Jianshu Zhang, Chengxuan Qian, Haosen Sun, Haoran Lu, Dingcheng Wang, Letian Xue, Han Liu. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Lifelong reinforcement learning (RL) has been developed as a paradigm for extending single-task RL to more realistic, dynamic settings. In lifelong RL, the "life" of an RL agent is modeled as a stream of tasks drawn from a task distribution. We propose EPIC (Empirical PAC-Bayes that Improves Continuously), a novel algorithm designed for lifelong RL using PAC-Bayes theory. EPIC learns a shared policy distribution, referred to as the world policy, which enables rapid adaptation to new tasks while retaining valuable knowledge from previous experiences. Our theoretical analysis establishes a relationship between the algorithm’s generalization performance and the number of prior tasks preserved in memory. We also derive the sample complexity of EPIC in terms of RL regret. Extensive experiments on a variety of environments demonstrate that EPIC significantly outperforms existing methods in lifelong RL, offering both theoretical guarantees and practical efficacy through the use of the world policy.

We study the computational limits of Low-Rank Adaptation (LoRA) for finetuning transformer-based models using fine-grained complexity theory. Our key observation is that the existence of low-rank decompositions within the gradient computation of LoRA adaptation leads to possible algorithmic speedup. This allows us to (i) identify a phase transition behavior of efficiency \blue{assuming the Strong Exponential Time Hypothesis (SETH)}, and (ii) prove the existence of almost linear algorithms by controlling the LoRA update computation term by term. For the former, we identify a sharp transition in the efficiency of all possible rank-$r$ LoRA update algorithms for transformers, based on specific norms resulting from the multiplications of the input sequence $X$, pretrained weights ${W^\star}$, and adapter matrices $\alpha B A/r$. Specifically, we derive a shared upper bound threshold for such norms and show that efficient (sub-quadratic) approximation algorithms of LoRA exist only below this threshold. For the latter, we prove the existence of almost linear approximation algorithms for LoRA adaptation by utilizing the hierarchical low-rank structures of LoRA gradients and approximating the gradients with a series of chained low-rank approximations. To showcase our theory, we consider two practical scenarios: partial (e.g., only $W_V$ and $W_Q$) and full adaptations (e.g., $W_Q$, $W_V$, and $W_K$) of weights in attention heads.

We investigate the statistical and computational limits of prompt tuning for transformer-based foundation models. Our key contributions are that prompt tuning on *single-head* transformers with only a *single* self-attention layer: (i) is universal, and (ii) supports efficient (even almost-linear time) algorithms under the Strong Exponential Time Hypothesis (SETH). Statistically, we prove that prompt tuning on such the simplest possible transformers are universal approximators for sequence-to-sequence Lipschitz functions. In addition, we provide an exponential-in-$dL$ and -in-$(1/\epsilon)$ lower bound on the required soft-prompt tokens for prompt tuning to memorize any dataset with 1-layer, 1-head transformers. Computationally, we identify a phase transition in the efficiency of prompt tuning, determined by the norm of the *soft-prompt-induced* keys and queries, and provide an upper bound criterion. Beyond this criterion, no sub-quadratic (efficient) algorithm for prompt tuning exists under SETH. Within this criterion, we showcase our theory by proving the existence of almost-linear time prompt tuning inference algorithms. These fundamental limits provide important necessary conditions for designing expressive and efficient prompt tuning methods for practitioners.

We investigate the approximation and estimation rates of conditional diffusion transformers (DiTs) with classifier-free guidance. We present a comprehensive analysis for “in-context” conditional DiTs under various common assumptions: generic and strong Hölder, linear latent (subspace), and Lipschitz score function assumptions. Importantly, we establish minimax optimality of DiTs by leveraging score function regularity. Specifically, we discretize the input domains into infinitesimal grids and then perform term-by-term Taylor expansions on the conditional diffusion score function under the Hölder smooth data assumption. This enables fine-grained use of transformers’ universal approximation through a more detailed piecewise constant approximation, and hence obtains tighter bounds. Additionally, we extend our analysis to latent settings. Our findings establish statistical limits for DiTs and offer practical guidance toward more efficient and accurate designs.

We present a Chain-of-Action (CoA) framework for multimodal and retrieval-augmented Question-Answering (QA). Compared to the literature, CoA overcomes two major challenges of current QA applications: (i) unfaithful hallucination that is inconsistent with real-time or domain facts and (ii) weak reasoning performance over compositional information. Our key contribution is a novel reasoning-retrieval mechanism that decomposes a complex question into a reasoning chain via systematic prompting and pre-designed actions. Methodologically, we propose three types of domain-adaptable `Plug-and-Play' actions for retrieving real-time information from heterogeneous sources. We also propose a multi-reference faith score to verify conflicts in the answers. In addition, our system demonstrates that detecting the knowledge boundaries of LLMs can significantly reduce both LLM interaction frequency and tokens usage in QA tasks. Empirically, we exploit both public benchmarks and a Web3 case study to demonstrate the capability of CoA over other methods.

We present a nonparametric interpretation for deep learning compatible modern Hopfield models and utilize this new perspective to debut efficient variants. Our key contribution stems from interpreting the memory storage and retrieval processes in modern Hopfield models as a nonparametric regression problem subject to a set of query-memory pairs. Interestingly, our framework not only recovers the known results from the original dense modern Hopfield model but also fills the void in the literature regarding efficient modern Hopfield models, by introducing sparse-structured modern Hopfield models with sub-quadratic complexity. We establish that this sparse model inherits the appealing theoretical properties of its dense analogue — connection with transformer attention, fixed point convergence and exponential memory capacity. Additionally, we showcase the versatility of our framework by constructing a family of modern Hopfield models as extensions, including linear, random masked, top-$K$ and positive random feature modern Hopfield models. Empirically, we validate our framework in both synthetic and realistic settings for memory retrieval and learning tasks.

To address the challenge of scarce computational resources in genomic modeling, we introduce GERM, a genomic foundation model optimized for accessibility and adaptability. GERM improves upon models like DNABERT-2 by eliminating outliers that hinder low-rank adaptation and post-training quantization, enhancing both efficiency and robustness. We replace the vanilla attention layer with an outlier-free mechanism inspired by associative memory models. By removing outliers during both pre-training and fine-tuning, this approach accelerates adaptation, reduces computational costs, and enhances quantization robustness within acceptable loss margins. Additionally, we propose GERM-T, a strategy that employs small-step continual learning within the outlier-free framework, leveraging original checkpoints to avoid retraining from scratch. Empirically, GERM improves fine-tuning performance by 37.98% and quantization by 64.34% over the baseline model. It also reduces average kurtosis by 92.14% and maximum infinity norm by 82.77%. Compared to leading methods, GERM consistently delivers superior performance, offering a practical solution for genomic modeling in resource-constrained settings.

We provide an exactly solvable example for interpreting In-Context Learning (ICL) with one-layer attention models as conditional retrieval of dense associative memory models. Our main contribution is to interpret ICL as memory reshaping in the modern Hopfield model from a conditional memory set (in-context examples). Specifically, we show that the in-context sequential examples induce an effective reshaping of the energy landscape of a Hopfield model. We integrate this in-context memory reshaping phenomenon into the existing Bayesian model averaging view of ICL [Zhang et al., AISTATS 2025] via the established equivalence between the modern Hopfield model and transformer attention. Under this unique perspective, we not only characterize how in-context examples shape predictions in the Gaussian linear regression case, but also recover the known $\epsilon$-stability generalization bound of the ICL for the one-layer attention model. We also give explanations for three key behaviors of ICL and validate them through experiments.

We investigate the transformer’s capability for in-context learning (ICL) to simulate the training process of deep models. Our key contribution is providing a positive example of using a transformer to train a deep neural network by gradient descent in an implicit fashion via ICL. Specifically, we provide an explicit construction of a $(2N+4)L$-layer transformer capable of simulating $L$ gradient descent steps of an $N$-layer ReLU network through ICL. We also give the theoretical guarantees for the approximation within any given error and the convergence of the ICL gradient descent. Additionally, we extend our analysis to the more practical setting using Softmax-based transformers. We validate our findings on synthetic datasets for 3-layer, 4-layer, and 6-layer neural networks. The results show that ICL performance matches that of direct training.

We establish the universal approximation capability of single-layer, single-head self- and cross-attention mechanisms with minimal attached structures. Our key insight is to interpret single-head attention as an input domain-partition mechanism that assigns distinct values to subregions. This allows us to engineer the attention weights such that this assignment imitates the target function. Building on this, we prove that a single self-attention layer, preceded by sum-of-linear transformations, is capable of approximating any continuous function on a compact domain under the $L_\infty$-norm. Furthermore, we extend this construction to approximate any Lebesgue integrable function under $L_p$-norm for $1\leq p <\infty$. Lastly, we also extend our techniques and show that, for the first time, single-head cross-attention achieves the same universal approximation guarantees.

Flow matching is an emerging generative modeling framework that learns continuous-time dynamics to map noise into data. To enhance expressiveness and sampling efficiency, recent works have explored incorporating high-order trajectory information. Despite the empirical success, a holistic theoretical foundation is still lacking. We present a unified framework for standard and high-order flow matching that incorporates trajectory derivatives up to an arbitrary order $K$. Our key innovation is establishing the marginalization technique that converts the intractable $K$-order loss into a simple conditional regression with exact gradients and identifying the consistency constraint. We establish sharp statistical rates of the $K$-order flow matching implemented with transformer networks. With $n$ samples, flow matching estimates nonparametric distributions at a rate $\tilde{O}(n^{-\Theta(1/d )})$, matching minimax lower bounds up to logarithmic factors.

Times series foundation models (TSFMs) have emerged as a promising paradigm for time series analyses and forecasting, showing remarkable generalization performance across different domains. Despite the efforts made on hallucinations of foundation models, hallucinations of TSFMs have been underexplored in existing literature. In this paper, we formally define TSFM hallucinations in the zero-shot forecasting setting by examining whether a generated forecast exhibits different dynamics from those of the context. Our study reveals that TSFM hallucinations are associated with the loss of context information in hidden states during forward propagation. As such, we propose a methodology to identify signal subspaces of TSFMs and magnify the information through intervention. Experiments demonstrate that our proposed intervention approach effectively mitigates hallucinations and improves forecasting performance. The signal strength measure computed from signal subspaces shows strong predictive power of hallucinations and forecasting performance of the model. Our work contributes to deeper understanding of TSFM trustworthiness that could foster future research in this direction.

SUMMARY: We introduce DNABERT-S, a tailored genome model that develops species-aware embeddings to naturally cluster and segregate DNA sequences of different species in the embedding space. Differentiating species from genomic sequences (i.e. DNA and RNA) is vital yet challenging, since many real-world species remain uncharacterized, lacking known genomes for reference. Embedding-based methods are therefore used to differentiate species in an unsupervised manner. DNABERT-S builds upon a pre-trained genome foundation model named DNABERT-2. To encourage effective embeddings to error-prone long-read DNA sequences, we introduce Manifold Instance Mixup (MI-Mix), a contrastive objective that mixes the hidden representations of DNA sequences at randomly selected layers and trains the model to recognize and differentiate these mixed proportions at the output layer. We further enhance it with the proposed Curriculum Contrastive Learning (C2LR) strategy. Empirical results on 28 diverse datasets show DNABERT-S's effectiveness, especially in realistic label-scarce scenarios. For example, it identifies twice more species from a mixture of unlabeled genomic sequences, doubles the Adjusted Rand Index (ARI) in species clustering, and outperforms the top baseline's performance in 10-shot species classification with just a 2-shot training. AVAILABILITY AND IMPLEMENTATION: Model, codes, and data are publically available at https://github.com/MAGICS-LAB/DNABERT_S.

Real-time dynamic scheduling in modern manufacturing is highly complex due to frequent disturbances and intricate operational constraints. While reinforcement learning (RL) has shown promise, existing approaches often rely on extensive dispatching rules and struggle to scale to factory-wide settings. This study introduces a scalable multi-agent RL (MARL) framework with a leader–follower structure, enabling decentralized agents to handle sub-problems while maintaining global coordination through abstract goals. To further enhance robustness, a limited rule-based conversion algorithm is proposed to mitigate performance degradation from poor agent decisions. Our experimental results demonstrate that the proposed model outperforms the state-of-the-art deep RL-based scheduling methods in various aspects. Additionally, the proposed model provides the most robust scheduling performance to demand changes. Overall, the proposed MARL-based scheduling model presents a promising solution to the real-time scheduling problem, with potential applications in various manufacturing industries.

Multivariate time series forecasting (MTSF) aims to predict future values of multiple variables based on past values of multivariate time series, and has been applied in fields including traffic flow prediction, stock price forecasting, and anomaly detection. Capturing the inter-dependencies among multiple series poses one significant challenge to MTSF. Recent works have considered modeling the correlated series as graph nodes and using graph neural network (GNN)-based approaches with attention mechanisms added to improve the test prediction accuracy, however, none of them have theoretical guarantees regarding the generalization error. In this paper, we develop a new norm-bounded graph attention network (GAT) for MTSF by upper-bounding the Frobenius norm of weights in each layer of the GAT model to enhance performance. We theoretically establish that the generalization error bound for our model is associated with various components of GAT models: the number of attention heads, the maximum number of neighbors, the upper bound of the Frobenius norm of the weight matrix in each layer, and the norm of the input features. Empirically, we investigate the impact of different components of GAT models on the generalization performance of MTSF on real data. Our experiment verifies our theoretical findings. We compare with multiple prior frequently cited graph-based methods for MTSF using real data sets and the experiment results show our method can achieve the best performance for MTSF. Our method provides novel perspectives for improving the generalization performance of MTSF, and our theoretical guarantees give substantial implications for designing graph-based methods with attention mechanisms for MTSF.

Decoding the linguistic intricacies of the genome is a crucial problem in biology, and pre-trained foundational models such as DNABERT and Nucleotide Transformer have made significant strides in this area. Existing works have largely hinged on k-mer, fixed-length permutations of A, T, C, and G, as the token of the genome language due to its simplicity. However, we argue that the computation and sample inefficiencies introduced by k-mer tokenization are primary obstacles in developing large genome foundational models. We provide conceptual and empirical insights into genome tokenization, building on which we propose to replace k-mer tokenization with Byte Pair Encoding (BPE), a statistics-based data compression algorithm that constructs tokens by iteratively merging the most frequent co-occurring genome segment in the corpus. We demonstrate that BPE not only overcomes the limitations of k-mer tokenization but also benefits from the computational efficiency of non-overlapping tokenization. Based on these insights, we introduce DNABERT-2, a refined genome foundation model that adapts an efficient tokenizer and employs multiple strategies to overcome input length constraints, reduce time and memory expenditure, and enhance model capability. Furthermore, we identify the absence of a comprehensive and standardized benchmark for genome understanding as another significant impediment to fair comparative analysis. In response, we propose the Genome Understanding Evaluation (GUE), a comprehensive multi-species genome classification dataset that amalgamates $36$ distinct datasets across $9$ tasks, with input lengths ranging from $70$ to $10000$. Through comprehensive experiments on the GUE benchmark, we demonstrate that DNABERT-2 achieves comparable performance to the state-of-the-art model with $21 \times$ fewer parameters and approximately $92 \times$ less GPU time in pre-training. Compared to DNABERT, while being $3 \times$ more efficient, DNABERT-2 outperforms it on $23$ out of $28$ datasets, with an average improvement of $6$ absolute scores on GUE. The code, data, and pre-trained model are available at \url{https://github.com/MAGICS-LAB/DNABERT_2}.

We present **STanHop-Net** (**S**parse **Tan**dem **Hop**field **Net**work) for multivariate time series prediction with memory-enhanced capabilities. At the heart of our approach is **STanHop**, a novel Hopfield-based neural network block, which sparsely learns and stores both temporal and cross-series representations in a data-dependent fashion. In essence, STanHop sequentially learns temporal representation and cross-series representation using two tandem sparse Hopfield layers. Additionally, STanHop incorporates two external memory modules: **Plug-and-Play** and **Tune-and-Play** for train-less and task-aware memory enhancements, respectively. They allow StanHop-Net to swiftly respond to sudden events. Methodologically, we construct the STanHop-Net by stacking STanHop blocks in a hierarchical fashion, enabling multi-resolution feature extraction with resolution-specific sparsity. Theoretically, we introduce a unified construction (**Generalized Sparse Modern Hopfield Model**) for both dense and sparse modern Hopfield models and show that it endows a tighter memory retrieval error compared to the dense counterpart without sacrificing memory capacity. Empirically, we validate the efficacy of STanHop-Net on many settings: time series prediction, fast test-time adaptation, and strongly correlated time series prediction.

We introduce an Outlier-Efficient Modern Hopfield Model (termed OutEffHop) and use it to address the outlier inefficiency problem of training gigantic transformer-based models. Our main contribution is a novel associative memory model facilitating outlier-efficient associative memory retrievals. Interestingly, this memory model manifests a model-based interpretation of an outlier-efficient attention mechanism (Softmax_1): it is an approximation of the memory retrieval process of OutEffHop. Methodologically, this allows us to introduce novel outlier-efficient Hopfield layers as powerful alternatives to traditional attention mechanisms, with superior post-quantization performance. Theoretically, the Outlier-Efficient Modern Hopfield Model retains and improves the desirable properties of standard modern Hopfield models, including fixed point convergence and exponential storage capacity. Empirically, we demonstrate the efficacy of the proposed model across large-scale transformer-based and Hopfield-based models (including BERT, OPT, ViT, and STanHop-Net), benchmarking against state-of-the-art methods like Clipped_Softmax and Gated_Attention. Notably, OutEffHop achieves an average reduction of 22+% in average kurtosis and 26+% in the maximum infinity norm of model outputs across four models. Code is available at GitHub; future updates are on arXiv.

We investigate the computational limits of the memory retrieval dynamics of modern Hopfield models from the fine-grained complexity analysis. Our key contribution is the characterization of a phase transition behavior in the efficiency of all possible modern Hopfield models based on the norm of patterns. Specifically, we establish an upper bound criterion for the norm of input query patterns and memory patterns. Only below this criterion, sub-quadratic (efficient) variants of the modern Hopfield model exist, assuming the Strong Exponential Time Hypothesis (SETH). To showcase our theory, we provide a formal example of efficient constructions of modern Hopfield models using low-rank approximation when the efficient criterion holds. This includes a derivation of a lower bound on the computational time, scaling linearly with $\max$\{ \# of stored memory patterns, length of input query sequence\}. In addition, we prove its memory retrieval error bound and exponential memory capacity.

We propose a two-stage optimization formulation for the memory retrieval dynamics of modern Hopfield models, termed $\mathtt{U\text{-}Hop}$. Our key contribution is a learnable feature map $\Phi$ which transforms the Hopfield energy function into a kernel space. This transformation ensures convergence between the local minima of energy and the fixed points of retrieval dynamics within the kernel space. Consequently, the kernel norm induced by $\Phi$ serves as a novel similarity measure. It utilizes the stored memory patterns as learning data to enhance memory capacity across all modern Hopfield models. Specifically, we accomplish this by constructing a separation loss $\mathcal{L}_\Phi$ that separates the local minima of kernelized energy by separating stored memory patterns in kernel space. Methodologically, $\mathtt{U\text{-}Hop}$ memory retrieval process consists of: **(Stage I:)** minimizing separation loss for a more uniformed memory (local minimum) distribution, followed by **(Stage II:)** standard Hopfield energy minimization for memory retrieval. This results in significant reduction of possible meta-stable states in the Hopfield energy function, thus preventing memory confusion. Empirically, with real-world datasets, we demonstrate that $\mathtt{U\text{-}Hop}$ outperforms all existing modern Hopfield models and SOTA similarity measures, achieving a substantial margin in both associative memory retrieval and deep learning tasks. Code is available at [GitHub](https://github.com/MAGICS-LAB/UHop); future updates are on [arXiv](https://arxiv.org/abs/2404.03827).

We introduce the **Bi**-Directional **S**parse **Hop**field Network (**BiSHop**), a novel end-to-end framework for tabular learning. BiSHop handles the two major challenges of deep tabular learning: non-rotationally invariant data structure and feature sparsity in tabular data. Our key motivation comes from the recently established connection between associative memory and attention mechanisms. Consequently, BiSHop uses a dual-component approach, sequentially processing data both column-wise and row-wise through two interconnected directional learning modules. Computationally, these modules house layers of generalized sparse modern Hopfield layers, a sparse extension of the modern Hopfield model with learnable sparsity. Methodologically, BiSHop facilitates multi-scale representation learning, capturing both intra-feature and inter-feature interactions, with adaptive sparsity at each scale. Empirically, through experiments on diverse real-world datasets, BiSHop surpasses current SOTA methods with significantly fewer HPO runs, marking it a robust solution for deep tabular learning. The code is available on [GitHub](https://github.com/MAGICS-LAB/BiSHop); future updates are on [arXiv](https://arxiv.org/abs/2404.03830).

We propose a new system named DOS®(Deployment Operating System for Robots) for reliably deploying any data-driven robots in both production and simulation environments. Compared to existing systems, DOS®features a unique CI/CD (continuous integration and continuous deployment) architecture which allows us to seamlessly integrate agile development and reliable operation in a fully automated fashion. With this CI/CD architecture, this paper mainly introduces three essential components that uniquely differentiate DOS®from existing robotic systems: (i) An environment adapter that provides a systematic and robust approach to handle the deployment complexity in real world environments; (ii) A data replay reservoir that provides a unified data model supporting arbitrary robotic decision models; (iii) An analytical profiler that collects any set of user-defined performance metrics for system optimization. DOS®significantly increases the reliability and maintainability of the deployed robotic systems. To illustrate this point, we compare DOS®with more traditional approaches on deploying a navigational robot in a challenging working environment with many new corner case scenarios. Our results show that DOS®outperforms traditional approach in great magnitudes in terms of deployment time and operational robustness.

Despite the widespread success of Transformers across various domains, their optimization guarantees in large-scale model settings are not well-understood. This paper rigorously analyzes the convergence properties of gradient flow in training Transformers with weight decay regularization. First, we construct the mean-field limit of large-scale Transformers, showing that as the model width and depth go to infinity, gradient flow converges to the Wasserstein gradient flow, which is represented by a partial differential equation. Then, we demonstrate that the gradient flow reaches a global minimum consistent with the PDE solution when the weight decay regularization parameter is sufficiently small. Our analysis is based on a series of novel mean-field techniques that adapt to Transformers. Compared with existing tools for deep networks (Lu et al., 2020) that demand homogeneity and global Lipschitz smoothness, we utilize a refined analysis assuming only $\textit{partial homogeneity}$ and $\textit{local Lipschitz smoothness}$. These new techniques may be of independent interest.

We study the optimal memorization capacity of modern Hopfield models and Kernelized Hopfield Models (KHMs), a transformer-compatible class of Dense Associative Memories. We present a tight analysis by establishing a connection between the memory configuration of KHMs and spherical codes from information theory. Specifically, we treat the stored memory set as a specialized spherical code. This enables us to cast the memorization problem in KHMs into a point arrangement problem on a hypersphere. We show that the optimal capacity of KHMs occurs when the feature space allows memories to form an optimal spherical code. This unique perspective leads to: 1. An analysis of how KHMs achieve optimal memory capacity, and identify corresponding necessary conditions. Importantly, we establish an upper capacity bound that matches the well-known exponential lower bound in the literature. This provides the first tight and optimal asymptotic memory capacity for modern Hopfield models. 2. A sub-linear time algorithm $\mathtt{U}\text{-}\mathtt{Hop}$+ to reach KHMs' optimal capacity. 3. An analysis of the scaling behavior of the required feature dimension relative to the number of stored memories. These efforts improve both the retrieval capability of KHMs and the representation learning of corresponding transformers. Experimentally, we provide thorough numerical results to back up theoretical findings.

We investigate the statistical and computational limits of latent **Di**ffusion **T**ransformers (**DiTs**) under the low-dimensional linear latent space assumption. Statistically, we study the universal approximation and sample complexity of the DiTs score function, as well as the distribution recovery property of the initial data. Specifically, under mild data assumptions, we derive an approximation error bound for the score network of latent DiTs, which is sub-linear in the latent space dimension. Additionally, we derive the corresponding sample complexity bound and show that the data distribution generated from the estimated score function converges toward a proximate area of the original one. Computationally, we characterize the hardness of both forward inference and backward computation of latent DiTs, assuming the Strong Exponential Time Hypothesis (SETH). For forward inference, we identify efficient criteria for all possible latent DiTs inference algorithms and showcase our theory by pushing the efficiency toward almost-linear time inference. For backward computation, we leverage the low-rank structure within the gradient computation of DiTs training for possible algorithmic speedup. Specifically, we show that such speedup achieves almost-linear time latent DiTs training by casting the DiTs gradient as a series of chained low-rank approximations with bounded error. Under the low-dimensional assumption, we show that the statistical rates and the computational efficiency are all dominated by the dimension of the subspace, suggesting that latent DiTs have the potential to bypass the challenges associated with the high dimensionality of initial data.

Transformers have achieved great success in recent years. Interestingly, transformers have shown particularly strong in-context learning capability -- even without fine-tuning, they are still able to solve unseen tasks well purely based on task-specific prompts. In this paper, we study the capability of one-layer transformers in learning the one-nearest neighbor prediction rule. Under a theoretical framework where the prompt contains a sequence of labeled training data and unlabeled test data, we show that, although the loss function is nonconvex, when trained with gradient descent, a single softmax attention layer can successfully learn to behave like a one-nearest neighbor classifier. Our result gives a concrete example on how transformers can be trained to implement nonparametric machine learning algorithms, and sheds light on the role of softmax attention in transformer models.

This paper addresses the challenges in accurate and real-time traffic congestion prediction under uncertainty by proposing Ising-Traffic, a dual-model Ising-based traffic prediction framework that delivers higher accuracy and lower latency than SOTA solutions. While traditional solutions face the dilemma from the trade-off between algorithm complexity and computational efficiency, our Ising-based method breaks away from the trade-off leveraging the Ising model's strong expressivity and the Ising machine's strong computation power. In particular, Ising-Traffic formulates traffic prediction under uncertainty into two Ising models: Reconstruct-Ising and Predict-Ising. Reconstruct-Ising is mapped onto modern Ising machines and handles uncertainty in traffic accurately with negligible latency and energy consumption, while Predict-Ising is mapped onto traditional processors and predicts future congestion precisely with only at most 1.8% computational demands of existing solutions. Our evaluation shows Ising-Traffic delivers on average 98X speedups and 5% accuracy improvement over SOTA.

We introduce the concept of programmable feature engineering for time series modeling and propose a feature programming framework. This framework generates large amounts of predictive features for noisy multivariate time series while allowing users to incorporate their inductive bias with minimal effort. The key motivation of our framework is to view any multivariate time series as a cumulative sum of fine-grained trajectory increments, with each increment governed by a novel spin-gas dynamical Ising model. This fine-grained perspective motivates the development of a parsimonious set of operators that summarize multivariate time series in an abstract fashion, serving as the foundation for large-scale automated feature engineering. Numerically, we validate the efficacy of our method on several synthetic and real-world noisy time series datasets.

We propose EMS®, a cloud-enabled massive computational experiment management system supporting high-throughput computational robotics research. Compared to existing systems, EMS® features a sky-based pipeline orchestrator which allows us to exploit heterogeneous computing environments painlessly (e.g., on-premise clusters, public clouds, edge devices) to optimally deploy large-scale computational jobs (e.g., with more than millions of computational hours) in an integrated fashion. Cornerstoned on this sky-based pipeline orchestrator, this paper introduces three abstraction layers of the EMS® software architecture: (i) Configuration management layer focusing on automatically enumerating experimental configurations; (ii) Dependency management layer focusing on managing the complex task dependencies within each experimental configuration; (iii) Computation management layer focusing on optimally executing the computational tasks using the given computing resource. Such an architectural design greatly increases the scalability and reproducibility of data-driven robotics research leading to much-improved productivity. To demonstrate this point, we compare EMS® with more traditional approaches on an offline reinforcement learning problem for training mobile robots. Our results show that EMS® outperforms more traditional approaches in two magnitudes of orders (in terms of experimental high throughput and cost) with only several lines of code change. We also exploit EMS® to develop mobile robot, robot arm, and bipedal applications, demonstrating its applicability to numerous robot applications.

We introduce the sparse modern Hopfield model as a sparse extension of the modern Hopfield model. Like its dense counterpart, the sparse modern Hopfield model equips a memory-retrieval dynamics whose one-step approximation corresponds to the sparse attention mechanism. Theoretically, our key contribution is a principled derivation of a closed-form sparse Hopfield energy using the convex conjugate of the sparse entropic regularizer. Building upon this, we derive the sparse memory retrieval dynamics from the sparse energy function and show its one-step approximation is equivalent to the sparse-structured attention. Importantly, we provide a sparsity-dependent memory retrieval error bound which is provably tighter than its dense analog. The conditions for the benefits of sparsity to arise are therefore identified and discussed. In addition, we show that the sparse modern Hopfield model maintains the robust theoretical properties of its dense counterpart, including rapid fixed point convergence and exponential memory capacity. Empirically, we use both synthetic and real-world datasets to demonstrate that the sparse Hopfield model outperforms its dense counterpart in many situations.

We propose efficient Langevin Monte Carlo algorithms for sampling distributions with nonsmooth convex composite potentials, which is the sum of a continuously differentiable function and a possibly nonsmooth function. We devise such algorithms leveraging recent advances in convex analysis and optimization methods involving Bregman divergences, namely the Bregman–Moreau envelopes and the Bregman proximity operators, and in the Langevin Monte Carlo algorithms reminiscent of mirror descent. The proposed algorithms extend existing Langevin Monte Carlo algorithms in two aspects—the ability to sample nonsmooth distributions with mirror descent-like algorithms, and the use of the more general Bregman–Moreau envelope in place of the Moreau envelope as a smooth approximation of the nonsmooth part of the potential. A particular case of the proposed scheme is reminiscent of the Bregman proximal gradient algorithm. The efficiency of the proposed methodology is illustrated with various sampling tasks at which existing Langevin Monte Carlo methods are known to perform poorly.

MOTIVATION: Deciphering the language of non-coding DNA is one of the fundamental problems in genome research. Gene regulatory code is highly complex due to the existence of polysemy and distant semantic relationship, which previous informatics methods often fail to capture especially in data-scarce scenarios. RESULTS: To address this challenge, we developed a novel pre-trained bidirectional encoder representation, named DNABERT, to capture global and transferrable understanding of genomic DNA sequences based on up and downstream nucleotide contexts. We compared DNABERT to the most widely used programs for genome-wide regulatory elements prediction and demonstrate its ease of use, accuracy and efficiency. We show that the single pre-trained transformers model can simultaneously achieve state-of-the-art performance on prediction of promoters, splice sites and transcription factor binding sites, after easy fine-tuning using small task-specific labeled data. Further, DNABERT enables direct visualization of nucleotide-level importance and semantic relationship within input sequences for better interpretability and accurate identification of conserved sequence motifs and functional genetic variant candidates. Finally, we demonstrate that pre-trained DNABERT with human genome can even be readily applied to other organisms with exceptional performance. We anticipate that the pre-trained DNABERT model can be fined tuned to many other sequence analyses tasks. AVAILABILITY AND IMPLEMENTATION: The source code, pretrained and finetuned model for DNABERT are available at GitHub (https://github.com/jerryji1993/DNABERT). SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

This paper studies large scatter matrix estimation for heavy tailed distributions. The contributions of this paper are twofold. First, we propose and advocate to use a new distribution family, the pair-elliptical, for modeling the high dimensional data. The pair-elliptical is more flexible and easier to check the goodness of fit compared to the elliptical. Secondly, built on the pair-elliptical family, we advocate using quantile-based statistics for estimating the scatter matrix. For this, we provide a family of quantile-based statistics. They outperform the existing ones for better balancing the efficiency and robustness. In particular, we show that the propose estimators have comparable performance to the moment-based counterparts under the Gaussian assumption. The method is also tuning-free compared to Catoni's M-estimator for covariance matrix estimation. We further apply the method to conduct a variety of statistical methods. The corresponding theoretical properties as well as numerical performances are provided.

In this paper, we explore the slot tagging with only a few labeled support sentences (a.k.a.few-shot).Few-shot slot tagging faces a unique challenge compared to the other fewshot classification problems as it calls for modeling the dependencies between labels.But it is hard to apply previously learned label dependencies to an unseen domain, due to the discrepancy of label sets.To tackle this, we introduce a collapsed dependency transfer mechanism into the conditional random field (CRF) to transfer abstract label dependency patterns as transition scores.In the few-shot setting, the emission score of CRF can be calculated as a word's similarity to the representation of each label.To calculate such similarity, we propose a Label-enhanced Task-Adaptive Projection Network (L-TapNet) based on the stateof-the-art few-shot classification model -Tap-Net, by leveraging label name semantics in representing labels.Experimental results show that our model significantly outperforms the strongest few-shot learning baseline by 14.64 F1 scores in the one-shot setting. 1

We exploit Markov games as a framework for collision-free navigation of human-centered robots. Unlike the classical methods which formulate robot navigation as a single-agent Markov decision process with a static environment, our framework of Markov games adopts a multi-agent formulation with one primary agent representing the robot and the remaining auxiliary agents form a dynamic or even competing environment. Such a framework allows us to develop a path-following type adversarial training strategy to learn a robust decentralized collision avoidance policy. Through thorough experiments on both simulated and real-world mobile robots, we show that the learnt policy outperforms the state-of-the-art algorithms in both sample complexity and runtime robustness.

The goal of noisy high-dimensional phase retrieval is to estimate an $s$-sparse parameter $\boldsymbol{\beta}^*\in \mathbb{R}^d$ from $n$ realizations of the model $Y = (\mathbf{X}^T \boldsymbol{\beta}^*)^2 + \varepsilon$. Based on this model, we propose a significant semi-parametric generalization called misspecified phase retrieval (MPR), in which $Y = f(\mathbf{X}^T \boldsymbol{\beta}^*, \varepsilon)$ with unknown $f$ and $\operatorname{Cov}(Y, (\mathbf{X}^T \boldsymbol{\beta}^*)^2) > 0$. For example, MPR encompasses $Y = h(|\mathbf{X}^T \boldsymbol{\beta}^*|) + \varepsilon$ with increasing $h$ as a special case. Despite the generality of the MPR model, it eludes the reach of most existing semi-parametric estimators. In this paper, we propose an estimation procedure, which consists of solving a cascade of two convex programs and provably recovers the direction of $\boldsymbol{\beta}^*$. Furthermore, we prove that our procedure is minimax optimal over the class of MPR models. Interestingly, our minimax analysis characterizes the statistical price of misspecifying the link function in phase retrieval models. Our theory is backed up by thorough numerical results.

We consider the estimation and inference of graphical models that characterize the dependency structure of high-dimensional tensor-valued data. To facilitate the estimation of the precision matrix corresponding to each way of the tensor, we assume the data follow a tensor normal distribution whose covariance has a Kronecker product structure. A critical challenge in the estimation and inference of this model is the fact that its penalized maximum likelihood estimation involves minimizing a non-convex objective function. To address it, this paper makes two contributions: (i) In spite of the non-convexity of this estimation problem, we prove that an alternating minimization algorithm, which iteratively estimates each sparse precision matrix while fixing the others, attains an estimator with an optimal statistical rate of convergence. (ii) We propose a de-biased statistical inference procedure for testing hypotheses on the true support of the sparse precision matrices, and employ it for testing a growing number of hypothesis with false discovery rate (FDR) control. The asymptotic normality of our test statistic and the consistency of FDR control procedure are established. Our theoretical results are backed up by thorough numerical studies and our real applications on neuroimaging studies of Autism spectrum disorder and users' advertising click analysis bring new scientific findings and business insights. The proposed methods are encoded into a publicly available R package Tlasso.

We consider the problem of multi-agent reinforcement learning (MARL) in video game AI, where the agents are located in a spatial grid-world environment and the number of agents varies both within and across episodes. The challenge is to flexibly control an arbitrary number of agents while achieving effective collaboration. Existing MARL methods usually suffer from the trade-off between these two considerations. To address the issue, we propose a novel architecture that learns a spatial joint representation of all the agents and outputs grid-wise actions. Each agent will be controlled independently by taking the action from the grid it occupies. By viewing the state information as a grid feature map, we employ a convolutional encoder-decoder as the policy network. This architecture naturally promotes agent communication because of the large receptive field provided by the stacked convolutional layers. Moreover, the spatially shared convolutional parameters enable fast parallel exploration that the experiences discovered by one agent can be immediately transferred to others. The proposed method can be conveniently integrated with general reinforcement learning algorithms, e.g., PPO and Q-learning. We demonstrate the effectiveness of the proposed method in extensive challenging multi-agent tasks in StarCraft II.

We describe a new library named picasso, which implements a unified framework of pathwise coordinate optimization for a variety of sparse learning problems (e.g., sparse linear regression, sparse logistic regression, sparse Poisson regression and scaled sparse linear regression) combined with efficient active set selection strategies. Besides, the library allows users to choose different sparsity-inducing regularizers, including the convex $\ell_1$, nonvoncex MCP and SCAD regularizers. The library is coded in \texttt{C++} and has user-friendly R and Python wrappers. Numerical experiments demonstrate that picasso can scale up to large problems efficiently.

Nonparametric estimation of multivariate functions is an important problem in statistical machine learning with many applications, ranging from nonparametric regression to nonparametric graphical models. Several authors have proposed to estimate multivariate functions under the smoothing spline analysis of variance (SSANOVA) framework, which assumes that the multivariate function can be decomposed into the summation of main effects, two-way interaction effects, and higher order interaction effects. However, existing methods are not scalable to the dimension of the random variables and the order of interactions. We propose a LAyer-wiSE leaRning strategy (LASER) to estimate multivariate functions under the SSANOVA framework. The main idea is to approximate the multivariate function sequentially starting from a model with only the main effects. Conditioned on the support of the estimated main effects, we estimate the two-way interaction effects only when the corresponding main effects are estimated to be non-zero. This process is continued until no more higher order interaction effects are identified. The proposed strategy provides a data-driven approach for estimating multivariate functions under the SSANOVA framework. Our proposal yields a sequence of estimators. To establish the theoretical properties of the sequence of estimators, we establish the notion of post-selection persistency. Extensive numerical studies are performed to evaluate the performance of our algorithm.

We propose a general theory for studying the landscape of nonconvex optimization with underlying symmetric structures for a class of machine learning problems (e.g., low-rank matrix factorization, phase retrieval, and deep linear neural networks). In particular, we characterize the locations of stationary points and the null space of Hessian matrices of the objective function via the lens of invariant groups. As a major motivating example, we apply the proposed general theory to characterize the global landscape of the nonconvex optimization in low-rank matrix factorization problem. We illustrate how the rotational symmetry group gives rise to infinitely many nonisolated strict saddle points and equivalent global minima of the objective function. By explicitly identifying all stationary points, we divide the entire parameter space into three regions: (R1) the region containing the neighborhoods of all strict saddle points where the objective has negative curvature; (R2) the region containing neighborhoods of all global minima, where the objective enjoys strong convexity along certain directions; and (R3) the complement of the above regions, where the gradient has sufficiently large magnitude. We further extend our result to the matrix sensing problem. Such global landscape implies that strong global convergence guarantees for popular iterative algorithms with arbitrary initial solutions.

This paper proposes a distributed privacy-preserving sparse PCA (DPS-PCA) algorithm that generates a minimax-optimal sparse PCA estimator under differential privacy constraints. In a distributed optimization framework, data providers can use this algorithm to collaboratively analyze the union of their data sets while limiting the disclosure of their private information. DPS-PCA can recover the leading eigenspace of the population covariance at a geometric convergence rate, and simultaneously achieves the optimal minimax statistical error for high-dimensional data. Our algorithm provides fine-tuned control over the tradeoff between estimation accuracy and privacy preservation. Numerical simulations demonstrate that DPS-PCA significantly outperforms other privacy-preserving PCA methods in terms of estimation accuracy and computational efficiency.

Inspired by recent successes of Monte-Carlo tree search (MCTS) in a number of artificial intelligence (AI) application domains, we propose a reinforcement learning (RL) technique that iteratively applies MCTS on batches of small, finite-horizon versions of the original infinite-horizon Markov decision process. The terminal condition of the finite-horizon problems, or the leaf-node evaluator of the decision tree generated by MCTS, is specified using a combination of an estimated value function and an estimated policy function. The recommendations generated by the MCTS procedure are then provided as feedback in order to refine, through classification and regression, the leaf-node evaluator for the next iteration. We provide the first sample complexity bounds for a tree search-based RL algorithm. In addition, we show that a deep neural network implementation of the technique can create a competitive AI agent for the popular multi-player online battle arena (MOBA) game King of Glory.

We study the hypothesis testing problem of inferring the existence of combinatorial structures in undirected graphical models. Although there exist extensive studies on the information-theoretic limits of this problem, it remains largely unexplored whether such limits can be attained by efficient algorithms. In this paper, we quantify the minimum computational complexity required to attain the information-theoretic limits based on an oracle computational model. We prove that, for testing common combinatorial structures, such as clique, nearest neighbor graph and perfect matching, against an empty graph, or large clique against small clique, the information-theoretic limits are provably unachievable by tractable algorithms in general. More importantly, we define structural quantities called the weak and strong edge densities, which offer deep insight into the existence of such computational-statistical tradeoffs. To the best of our knowledge, our characterization is the first to identify and explain the fundamental tradeoffs between statistics and computation for combinatorial inference problems in undirected graphical models.

We consider the problem of learning high-dimensional Gaussian graphical models. The graphical lasso is one of the most popular methods for estimating Gaussian graphical models. However, it does not achieve the oracle rate of convergence. In this paper, we propose the graphical nonconvex optimization for optimal estimation in Gaussian graphical models, which is then approximated by a sequence of convex programs. Our proposal is computationally tractable and produces an estimator that achieves the oracle rate of convergence. The statistical error introduced by the sequential approximation using a sequence of convex programs is clearly demonstrated via a contraction property. The proposed methodology is then extended to modeling semiparametric graphical models. We show via numerical studies that the proposed estimator outperforms other popular methods for estimating Gaussian graphical models.

We consider the fully decentralized multi-agent reinforcement learning (MARL) problem, where the agents are connected via a time-varying and possibly sparse communication network. Specifically, we assume that the reward functions of the agents might correspond to different tasks, and are only known to the corresponding agent. Moreover, each agent makes individual decisions based on both the information observed locally and the messages received from its neighbors over the network. To maximize the globally averaged return over the network, we propose two fully decentralized actor-critic algorithms, which are applicable to large-scale MARL problems in an online fashion. Convergence guarantees are provided when the value functions are approximated within the class of linear functions. Our work appears to be the first theoretical study of fully decentralized MARL algorithms for networked agents that use function approximation.

We present computationally efficient algorithms to test various combinatorial structures of large-scale graphical models. In order to test the hypotheses on their topological structures, we propose two adjacency matrix sketching frameworks: neighborhood sketching and subgraph sketching. The neighborhood sketching algorithm is proposed to test the connectivity of graphical models. This algorithm randomly subsamples vertices and conducts neighborhood regression and screening. The global sketching algorithm is proposed to test the topological properties requiring exponential computation complexity, especially testing the chromatic number and the maximum clique. This algorithm infers the corresponding property based on the sampled subgraph. Our algorithms are shown to substantially accelerate the computation of existing methods. We validate our theory and method through both synthetic simulations and a real application in neuroscience.

We consider deep policy learning with only batched historical trajectories. The main challenge of this problem is that the learner no longer has a simulator or ``environment oracle'' as in most reinforcement learning settings. To solve this problem, we propose a monotonic advantage reweighted imitation learning strategy that is applicable to problems with complex nonlinear function approximation and works well with hybrid (discrete and continuous) action space. The method does not rely on the knowledge of the behavior policy, thus can be used to learn from data generated by an unknown policy. Under mild conditions, our algorithm, though surprisingly simple, has a policy improvement bound and outperforms most competing methods empirically. Thorough numerical results are also provided to demonstrate the efficacy of the proposed methodology.

Motivation: In recent years there have been several efforts to generate sensitivity profiles of collections of genomically characterized cell lines to panels of candidate therapeutic compounds. These data provide the basis for the development of in silico models of sensitivity based on cellular, genetic, or expression biomarkers of cancer cells. However, a remaining challenge is an efficient way to identify accurate sets of biomarkers to validate. To address this challenge, we developed methodology using gene-expression profiles of human cancer cell lines to predict the responses of these cell lines to a panel of compounds. Results: We developed an iterative weighting scheme which, when applied to elastic net, a regularized regression method, significantly improves the overall accuracy of predictions, particularly in the highly sensitive response region. In addition to application of these methods to actual chemical sensitivity data, we investigated the effects of sample size, number of features, model sparsity, signal-to-noise ratio, and feature correlation on predictive performance using a simulation framework, particularly for situations where the number of covariates is much larger than sample size. While our method aims to be useful in therapeutic discovery and understanding of the basic mechanisms of action of drugs and their targets, it is generally applicable in any domain where predictions of extreme responses are of highest importance. Availability and implementation: The iterative and other weighting algorithms were implemented in R. The code is available at https://github.com/kiwtir/RWEN. The CTRP data are available at ftp://caftpd.nci.nih.gov/pub/OCG-DCC/CTD2/Broad/CTRPv2.1_2016_pub_NatChemBiol_12_109/ and the Sanger data at ftp://ftp.sanger.ac.uk/pub/project/cancerrxgene/releases/release-6.0/. Supplementary information: Supplementary data are available at Bioinformatics online.

We propose a new class of semiparametric exponential family graphical models for the analysis of high dimensional mixed data. Different from the existing mixed graphical models, we allow the nodewise conditional distributions to be semiparametric generalized linear models with unspecified base measure functions. Thus, one advantage of our method is that it is unnecessary to specify the type of each node and the method is more convenient to apply in practice. Under the proposed model, we consider both problems of parameter estimation and hypothesis testing in high dimensions. In particular, we propose a symmetric pairwise score test for the presence of a single edge in the graph. Compared to the existing methods for hypothesis tests, our approach takes into account of the symmetry of the parameters, such that the inferential results are invariant with respect to the different parametrizations of the same edge. Thorough numerical simulations and a real data example are provided to back up our theoretical results.

We consider estimating the parametric component of single index models in high dimensions. Compared with existing work, we do not require the covariate to be normally distributed. Utilizing Stein’s Lemma, we propose estimators based on the score function of the covariate. Moreover, to handle score function and response variables that are heavy-tailed, our estimators are constructed via carefully thresholding their empirical counterparts. Under a bounded fourth moment condition, we establish optimal statistical rates of convergence for the proposed estimators. Extensive numerical experiments are provided to back up our theory.

High dimensional sparse learning has imposed a great computational challenge to large scale data analysis. In this paper, we investiage a broad class of sparse learning approaches formulated as linear programs parametrized by a {\em regularization factor}, and solve them by the parametric simplex method (PSM). PSM offers significant advantages over other competing methods: (1) PSM naturally obtains the complete solution path for all values of the regularization parameter; (2) PSM provides a high precision dual certificate stopping criterion; (3) PSM yields sparse solutions through very few iterations, and the solution sparsity significantly reduces the computational cost per iteration. Particularly, we demonstrate the superiority of PSM over various sparse learning approaches, including Dantzig selector for sparse linear regression, sparse support vector machine for sparse linear classification, and sparse differential network estimation. We then provide sufficient conditions under which PSM always outputs sparse solutions such that its computational performance can be significantly boosted. Thorough numerical experiments are provided to demonstrate the outstanding performance of the PSM method.

We consider estimating the parametric components of semiparametric multi-index models in high dimensions. To bypass the requirements of Gaussianity or elliptical symmetry of covariates in existing methods, we propose to leverage a second-order Stein’s method with score function-based corrections. We prove that our estimator achieves a near-optimal statistical rate of convergence even when the score function or the response variable is heavy-tailed. To establish the key concentration results, we develop a data-driven truncation argument that may be of independent interest. We supplement our theoretical findings with simulations.

In this paper, we present a nonconvex alternating minimization optimization algorithm for low-rank and sparse structure pursuit. Compared with convex relaxation based methods, the proposed algorithm is computationally more efficient for large scale problems. In our study, we define a notion of bounded difference of gradients, based on which we rigorously prove that with suitable initialization, the proposed nonconvex optimization algorithm enjoys linear convergence to the global optima and exactly recovers the underlying low rank and sparse matrices under standard conditions such as incoherence and sparsity conditions. For a wide range of statistical models such as multi-task learning and robust principal component analysis (RPCA), our algorithm provides a principled approach to learning the low rank and sparse structures with provable guarantee. Thorough experiments on both synthetic and real datasets backup our theory.

We propose a sequential learning policy for noisy discrete global optimization and ranking and selection (R&S) problems with high dimensional sparse belief functions, where there are hundreds or even thousands of features, but only a small portion of these features contain explanatory power. Our problem setting, motivated by the experimental sciences, arises where we have to choose which experiment to run next. Here the experiments are time-consuming and expensive. We derive a sparse knowledge gradient (SpKG) decision policy based on the \ell_1-penalized regression Lasso to identify the sparsity pattern before our budget is exhausted. This policy is a unique and novel hybrid of Bayesian R&S with a frequentist learning approach. Theoretically, we provide the error bound of the posterior mean estimate, which has shown to be at the minimax optimal \sqrts\log p/n rate. Controlled experiments on both synthetic data and real application for automatically designing experiments to identify the structure of an RNA molecule show that the algorithm efficiently learns the correct set of nonzero parameters. It also outperforms several other learning policies.

The cyclic block coordinate descent-type (CBCD-type) methods have shown remarkable computational performance for solving strongly convex minimization problems. Typical applications include many popular statistical machine learning methods such as elastic-net regression, ridge penalized logistic regression, and sparse additive regression. Existing optimization literature has shown that the CBCD-type methods attain iteration complexity of O(p⋅\log(1/ε)), where εis a pre-specified accuracy of the objective value, and p is the number of blocks. However, such iteration complexity explicitly depends on p, and therefore is at least p times worse than those of gradient descent methods. To bridge this theoretical gap, we propose an improved convergence analysis for the CBCD-type methods. In particular, we first show that for a family of quadratic minimization problems, the iteration complexity of the CBCD-type methods matches that of the GD methods in term of dependency on p (up to a \log^2 p factor). Thus our complexity bounds are sharper than the existing bounds by at least a factor of p/\log^2p. We also provide a lower bound to confirm that our improved complexity bounds are tight (up to a \log^2 p factor) if the largest and smallest eigenvalues of the Hessian matrix do not scale with p. Finally, we generalize our analysis to other strongly convex minimization problems beyond quadratic ones.

We propose a stochastic variance reduced optimization algorithm for solving a class of large-scale nonconvex optimization problems with cardinality constraints, and provide sufficient conditions under which the proposed algorithm enjoys strong linear convergence guarantees and optimal estimation accuracy in high dimensions. Numerical experiments demonstrate the efficiency of our method in terms of both parameter estimation and computational performance.

Many high dimensional sparse learning problems are formulated as nonconvex optimization. A popular approach to solve these nonconvex optimization problems is through convex relaxations such as linear and semidefinite programming. In this paper, we study the statistical limits of convex relaxations. Particularly, we consider two problems: Mean estimation for sparse principal submatrix and edge probability estimation for stochastic block model. We exploit the sum-of-squares relaxation hierarchy to sharply characterize the limits of a broad class of convex relaxations. Our result shows statistical optimality needs to be compromised for achieving computational tractability using convex relaxations. Compared with existing results on computational lower bounds for statistical problems, which consider general polynomial-time algorithms and rely on computational hardness hypotheses on problems like planted clique detection, our theory focuses on a broad class of convex relaxations and does not rely on unproven hypotheses.

We study parameter estimation for sparse nonlinear regression. More specifically, we assume the data are given by y = f( \bf x^T \bf β^* ) + ε, where f is nonlinear. To recover \bf βs, we propose an \ell_1-regularized least-squares estimator. Unlike classical linear regression, the corresponding optimization problem is nonconvex because of the nonlinearity of f. In spite of the nonconvexity, we prove that under mild conditions, every stationary point of the objective enjoys an optimal statistical rate of convergence. Detailed numerical results are provided to back up our theory.

In the information age, people can easily collect information about the same set of entities from multiple sources, among which conflicts are inevitable. This leads to an important task, truth discovery, i.e., to identify true facts (truths) via iteratively updating truths and source reliability. However, the convergence to the truths is never discussed in existing work, and thus there is no theoretical guarantee in the results of these truth discovery approaches. In contrast, in this paper we propose a truth discovery approach with theoretical guarantee. We propose a randomized gaussian mixture model (RGMM) to represent multi-source data, where truths are model parameters. We incorporate source bias which captures its reliability degree into RGMM formulation. The truth discovery task is then modeled as seeking the maximum likelihood estimate (MLE) of the truths. Based on expectation-maximization (EM) techniques, we propose population-based (i.e., on the limit of infinite data) and sample-based (i.e., on a finite set of samples) solutions for the MLE. Theoretically, we prove that both solutions are contractive to an ε-ball around the MLE, under certain conditions. Experimentally, we evaluate our method on both simulated and real-world datasets. Experimental results show that our method achieves high accuracy in identifying truths with convergence guarantee.

The importance of studying the robustness of learners to malicious data is well established. While much work has been done establishing both robust estimators and effective data injection attacks when the attacker is omniscient, the ability of an attacker to provably harm learning while having access to little information is largely unstudied. We study the potential of a “blind attacker” to provably limit a learner’s performance by data injection attack without observing the learner’s training set or any parameter of the distribution from which it is drawn. We provide examples of simple yet effective attacks in two settings: firstly, where an “informed learner” knows the strategy chosen by the attacker, and secondly, where a “blind learner” knows only the proportion of malicious data and some family to which the malicious distribution chosen by the attacker belongs. For each attack, we analyze minimax rates of convergence and establish lower bounds on the learner’s minimax risk, exhibiting limits on a learner’s ability to learn under data injection attack even when the attacker is “blind”.

Solving statistical learning problems often involves nonconvex optimization. Despite the empirical success of nonconvex statistical optimization methods, their global dynamics, especially convergence to the desirable local minima, remain less well understood in theory. In this paper, we propose a new analytic paradigm based on diffusion processes to characterize the global dynamics of nonconvex statistical optimization. As a concrete example, we study stochastic gradient descent (SGD) for the tensor decomposition formulation of independent component analysis. In particular, we cast different phases of SGD into diffusion processes, i.e., solutions to stochastic differential equations. Initialized from an unstable equilibrium, the global dynamics of SGD transit over three consecutive phases: (i) an unstable Ornstein-Uhlenbeck process slowly departing from the initialization, (ii) the solution to an ordinary differential equation, which quickly evolves towards the desirable local minimum, and (iii) a stable Ornstein-Uhlenbeck process oscillating around the desirable local minimum. Our proof techniques are based upon Stroock and Varadhan’s weak convergence of Markov chains to diffusion processes, which are of independent interest.

The goal of noisy high-dimensional phase retrieval is to estimate an $s$-sparse parameter $\boldsymbol{\beta}^*\in \mathbb{R}^d$ from $n$ realizations of the model $Y = (\boldsymbol{X}^{\top} \boldsymbol{\beta}^*)^2 + \varepsilon$. Based on this model, we propose a significant semi-parametric generalization called misspecified phase retrieval (MPR), in which $Y = f(\boldsymbol{X}^{\top}\boldsymbol{\beta}^*, \varepsilon)$ with unknown $f$ and $\operatorname{Cov}(Y, (\boldsymbol{X}^{\top}\boldsymbol{\beta}^*)^2) > 0$. For example, MPR encompasses $Y = h(|\boldsymbol{X}^{\top} \boldsymbol{\beta}^*|) + \varepsilon$ with increasing $h$ as a special case. Despite the generality of the MPR model, it eludes the reach of most existing semi-parametric estimators. In this paper, we propose an estimation procedure, which consists of solving a cascade of two convex programs and provably recovers the direction of $\boldsymbol{\beta}^*$. Our theory is backed up by thorough numerical results.

We consider the weakly supervised binary classification problem where the labels are randomly flipped with probability $1-\alpha$. Although there exist numerous algorithms for this problem, it remains theoretically unexplored how the statistical accuracies and computational efficiency of these algorithms depend on the degree of supervision, which is quantified by $\alpha$. In this paper, we characterize the effect of $\alpha$ by establishing the information-theoretic and computational boundaries, namely, the minimax-optimal statistical accuracy that can be achieved by all algorithms, and polynomial-time algorithms under an oracle computational model. For small $\alpha$, our result shows a gap between these two boundaries, which represents the computational price of achieving the information-theoretic boundary due to the lack of supervision. Interestingly, we also show that this gap narrows as $\alpha$ increases. In other words, having more supervision, i.e., more correct labels, not only improves the optimal statistical accuracy as expected, but also enhances the computational efficiency for achieving such accuracy.

Gaussian vector autoregressive (VAR) processes have been extensively studied in the literature. However, Gaussian assumptions are stringent for heavy-tailed time series that frequently arises in finance and economics. In this paper, we develop a unified framework for modeling and estimating heavy-tailed VAR processes. In particular, we generalize the Gaussian VAR model by an elliptical VAR model that naturally accommodates heavy-tailed time series. Under this model, we develop a quantile-based robust estimator for the transition matrix of the VAR process. We show that the proposed estimator achieves parametric rates of convergence in high dimensions. This is the first work in analyzing heavy-tailed high dimensional VAR processes. As an application of the proposed framework, we investigate Granger causality in the elliptical VAR process, and show that the robust transition matrix estimator induces sign-consistent estimators of Granger causality. The empirical performance of the proposed methodology is demonstrated by both synthetic and real data. We show that the proposed estimator is robust to heavy tails, and exhibit superior performance in stock price prediction.

We propose a robust portfolio optimization approach based on quantile statistics. The proposed method is robust to extreme events in asset returns, and accommodates large portfolios under limited historical data. Specifically, we show that the risk of the estimated portfolio converges to the oracle optimal risk with parametric rate under weakly dependent asset returns. The theory does not rely on higher order moment assumptions, thus allowing for heavy-tailed asset returns. Moreover, the rate of convergence quantifies that the size of the portfolio under management is allowed to scale exponentially with the sample size of the historical data. The empirical effectiveness of the proposed method is demonstrated under both synthetic and real stock data. Our work extends existing ones by achieving robustness in high dimensions, and by allowing serial dependence.

We consider the estimation of sparse graphical models that characterize the dependency structure of high-dimensional tensor-valued data. To facilitate the estimation of the precision matrix corresponding to each way of the tensor, we assume the data follow a tensor normal distribution whose covariance has a Kronecker product structure. The penalized maximum likelihood estimation of this model involves minimizing a non-convex objective function. In spite of the non-convexity of this estimation problem, we prove that an alternating minimization algorithm, which iteratively estimates each sparse precision matrix while fixing the others, attains an estimator with the optimal statistical rate of convergence as well as consistent graph recovery. Notably, such an estimator achieves estimation consistency with only one tensor sample, which is unobserved in previous work. Our theoretical results are backed by thorough numerical studies.

Abstract We propose a family of non-uniform sampling strategies to provably speed up a class of stochastic optimization algorithms with linear convergence including Stochastic Variance Reduced Gradient (SVRG) and Stochastic Dual Coordinate Ascent (SDCA). For a large family of penalized empirical risk minimization problems, our methods exploit data dependent local smoothness of the loss functions near the optimum, while maintaining convergence guarantees. Our bounds are the first to quantify the advantage gained from local smoothness which are significant for some problems significantly better. Empirically, we provide thorough numerical results to back up our theory. Additionally we present algorithms exploiting local smoothness in more aggressive ways, which perform even better in practice.

We provide a general theory of the expectation-maximization (EM) algorithm for inferring high dimensional latent variable models. In particular, we make two contributions: (i) For parameter estimation, we propose a novel high dimensional EM algorithm which naturally incorporates sparsity structure into parameter estimation. With an appropriate initialization, this algorithm converges at a geometric rate and attains an estimator with the (near-)optimal statistical rate of convergence. (ii) Based on the obtained estimator, we propose a new inferential procedure for testing hypotheses for low dimensional components of high dimensional parameters. For a broad family of statistical models, our framework establishes the first computationally feasible approach for optimal estimation and asymptotic inference in high dimensions.

Linear regression studies the problem of estimating a model parameter $\beta^* \in \R^p$, from $n$ observations $\{(y_i,x_i)\}_{i=1}^n$ from linear model $y_i = \langle \x_i,\beta^* \rangle + \epsilon_i$. We consider a significant generalization in which the relationship between $\langle x_i,\beta^* \rangle$ and $y_i$ is noisy, quantized to a single bit, potentially nonlinear, noninvertible, as well as unknown. This model is known as the single-index model in statistics, and, among other things, it represents a significant generalization of one-bit compressed sensing. We propose a novel spectral-based estimation procedure and show that we can recover $\beta^*$ in settings (i.e., classes of link function $f$) where previous algorithms fail. In general, our algorithm requires only very mild restrictions on the (unknown) functional relationship between $y_i$ and $\langle x_i,\beta^* \rangle$. We also consider the high dimensional setting where $\beta^*$ is sparse, and introduce a two-stage nonconvex framework that addresses estimation challenges in high dimensional regimes where $p \gg n$. For a broad class of link functions between $\langle x_i,\beta^* \rangle$ and $y_i$, we establish minimax lower bounds that demonstrate the optimality of our estimators in both the classical and high dimensional regimes.

We study the estimation of low rank matrices via nonconvex optimization. Compared with convex relaxation, nonconvex optimization exhibits superior empirical performance for large scale instances of low rank matrix estimation. However, the understanding of its theoretical guarantees are limited. In this paper, we define the notion of projected oracle divergence based on which we establish sufficient conditions for the success of nonconvex optimization. We illustrate the consequences of this general framework for matrix sensing and completion. In particular, we prove that a broad class of nonconvex optimization algorithms, including alternating minimization and gradient-type methods, geometrically converge to the global optimum and exactly recover the true low rank matrices under standard conditions.

UNLABELLED: One central theme of modern high-throughput genomic data analysis is to identify relevant genomic features as well as build up a predictive model based on selected features for various tasks such as personalized medicine. Correlating the large number of 'omics' features with a certain phenotype is particularly challenging due to small sample size (n) and high dimensionality (p). To address this small n, large p problem, various forms of sparse regression models have been proposed by exploiting the sparsity assumption. Among these, network-constrained sparse regression model is of particular interest due to its ability to utilize the prior graph/network structure in the omics data. Despite its potential usefulness for omics data analysis, no efficient R implementation is publicly available. Here we present an R software package 'glmgraph' that implements the graph-constrained regularization for both sparse linear regression and sparse logistic regression. We implement both the L1 penalty and minimax concave penalty for variable selection and Laplacian penalty for coefficient smoothing. Efficient coordinate descent algorithm is used to solve the optimization problem. We demonstrate the use of the package by applying it to a human microbiome dataset, where phylogeny structure among bacterial taxa is available. AVAILABILITY AND IMPLEMENTATION: 'glmgraph' is implemented in R and C++ Armadillo and publicly available under CRAN.

We consider the high-dimensional discriminant analysis problem. For this problem, different methods have been proposed and justified by establishing exact convergence rates for the classification risk, as well as the ℓ2convergence results to the discriminative rule. However, sharp theoretical analysis for the variable selection performance of these procedures have not been established, even though model interpretation is of fundamental importance in scientific data analysis. This paper bridges the gap by providing sharp sufficient conditions for consistent variable selection using the sparse discriminant analysis. Through careful analysis, we establish rates of convergence that are significantly faster than the best known results and admit an optimal scaling of the sample size n, dimensionality p, and sparsity level s in the high-dimensional setting. Sufficient conditions are complemented by the necessary information theoretic limits on the variable selection problem in the context of high-dimensional discriminant analysis. Exploiting a numerical equivalence result, our method also establish the optimal results for the ROAD estimator and the sparse optimal scoring estimator. Furthermore, we analyze an exhaustive search procedure, whose performance serves as a benchmark, and show that it is variable selection consistent under weaker conditions. Extensive simulations demonstrating the sharpness of the bounds are also provided.

Recent genomic studies have identified genes related to specific phenotypes. In addition to marginal association analysis for individual genes, analyzing gene pathways (functionally related sets of genes) may yield additional valuable insights. We have devised an approach to phenotype classification from gene expression profiling. Our method named “group Nearest Shrunken Centroids (gNSC)” is an enhancement of the Nearest Shrunken Centroids (NSC) which is a popular and scalable method to analyze big data. While fully utilizing the variable structure of gene pathways, gNSC shares comparable computational speed as NSC if the group size is small. Comparing with NSC, gNSC improves the power of classification by utilizing the gene pathway information. In practice, we investigate the performance of gNSC on one of the largest microarray datasets aggregated from the internet. We show the effectiveness of our method by comparing the misclassification rate of gNSC with that of NSC. Additionally, we present a novel application of NSC/gNSC on context analysis of association between pathways and certain medical words. Some newest biological findings are rediscovered.

We propose a new high dimensional semiparametric principal component analysis (PCA) method, named Copula Component Analysis (COCA). The semiparametric model assumes that, after unspecified marginally monotone transformations, the distributions are multivariate Gaussian. COCA improves upon PCA and sparse PCA in three aspects: (i) It is robust to modeling assumptions; (ii) It is robust to outliers and data contamination; (iii) It is scale-invariant and yields more interpretable results. We prove that the COCA estimators obtain fast estimation rates and are feature selection consistent when the dimension is nearly exponentially large relative to the sample size. Careful experiments confirm that COCA outperforms sparse PCA on both synthetic and real-world data sets.

We propose a semiparametric method for estimating a precision matrix of high-dimensional elliptical distributions. Unlike most existing methods, our method naturally handles heavy tailness and conducts parameter estimation under a calibration framework, thus achieves improved theoretical rates of convergence and finite sample performance on heavy-tail applications. We further demonstrate the performance of the proposed method using thorough numerical experiments.

We study sparse principal component analysis (sparse PCA) for high dimensional multivariate vector autoregressive (VAR) time series. By treating the transition matrix as a nuisance parameter, we show that sparse PCA can be directly applied on analyzing multivariate time series as if the data are i.i.d. generated. Under a double asymptotic framework in which both the length of the sample period T and dimensionality d of the time series can increase (with possibly d≫T), we provide explicit rates of convergence of the angle between the estimated and population leading eigenvectors of the time series covariance matrix. Our results suggest that the spectral norm of the transition matrix plays a pivotal role in determining the final rates of convergence. Implications of such a general result is further illustrated using concrete examples. The results of this paper have impacts on different applications, including financial time series, biomedical imaging, and social media, etc.

In this paper, we propose a new method in estimating transition matrices of high dimensional vector autoregressive (VAR) models. Here the data are assumed to come from a stationary Gaussian VAR time series. By formulating the problem as a linear program, we provide a new approach to conduct inference on such models. In theory, under a doubly asymptotic framework in which both the sample size T and dimensionality d of the time series can increase, we provide explicit rates of convergence between the estimator and the population transition matrix under different matrix norms. Our results show that the spectral norm of the transition matrix plays a pivotal role in determining the final rates of convergence. This is the first work analyzing the estimation of transition matrices under a high dimensional doubly asymptotic framework. Experiments are conducted on both synthetic and real-world stock data to demonstrate the effectiveness of the proposed method compared with the existing methods. The results of this paper have broad impact on different applications, including finance, genomics, and brain imaging.

In this paper, we analyze the performance of a semiparametric principal component analysis named Copula Component Analysis (COCA) (Han & Liu, 2012) when the data are dependent. The semiparametric model assumes that, after unspecified marginally monotone transformations, the distributions are multivariate Gaussian. We study the scenario where the observations are drawn from non-i.i.d. processes ($m$-dependency or a more general $\phi$-mixing case). We show that COCA can allow weak dependence. In particular, we provide the generalization bounds of convergence for both support recovery and parameter estimation of COCA for the dependent data. We provide explicit sufficient conditions on the degree of dependence, under which the parametric rate can be maintained. To our knowledge, this is the first work analyzing the theoretical performance of PCA for the dependent data in high dimensional settings. Our results strictly generalize the analysis in Han & Liu (2012) and the techniques we used have the separate interest for analyzing a variety of other multivariate statistical methods.

High-dimensional discriminant analysis is of fundamental importance in multivariate statistics. Existing theoretical results sharply characterize different procedures, providing sharp convergence results for the classification risk, as well as the l2 convergence results to the discriminative rule. However, sharp theoretical results for the problem of variable selection have not been established, even though model interpretation is of importance in many scientific domains. In this paper, we bridge this gap by providing sharp sufficient conditions for consistent variable selection using the ROAD estimator (Fan et al., 2010). Our results provide novel theoretical insights for the ROAD estimator. Sufficient conditions are complemented by the necessary information theoretic limits on variable selection in high-dimensional discriminant analysis. This complementary result also establishes optimality of the ROAD estimator for a certain family of problems.

Many real world network problems often concern multivariate nodal attributes such as image, textual, and multi-view feature vectors on nodes, rather than simple univariate nodal attributes. The existing graph estimation methods built on Gaussian graphical models and covariance selection algorithms can not handle such data, neither can the theories developed around such methods be directly applied. In this paper, we propose a new principled framework for estimating multi-attribute graphs. Instead of estimating the partial correlation as in current literature, our method estimates the partial canonical correlations that naturally accommodate complex nodal features. Computationally, we provide an efficient algorithm which utilizes the multi-attribute structure. Theoretically, we provide sufficient conditions which guarantee consistent graph recovery. Extensive simulation studies demonstrate performance of our method under various conditions.

In this paper we focus on the principal component regression and its application to high dimension non-Gaussian data. The major contributions are in two folds. First, in low dimensions and under a double asymptotic framework where both the dimension $d$ and sample size $n$ can increase, by borrowing the strength from recent development in minimax optimal principal component estimation, we first time sharply characterize the potential advantage of classical principal component regression over least square estimation under the Gaussian model. Secondly, we propose and analyze a new robust sparse principal component regression on high dimensional elliptically distributed data. The elliptical distribution is a semiparametric generalization of the Gaussian, including many well known distributions such as multivariate Gaussian, rank-deficient Gaussian, $t$, Cauchy, and logistic. It allows the random vector to be heavy tailed and have tail dependence. These extra flexibilities make it very suitable for modeling finance and biomedical imaging data. Under the elliptical model, we prove that our method can estimate the regression coefficients in the optimal parametric rate and therefore is a good alternative to the Gaussian based methods. Experiments on synthetic and real world data are conducted to illustrate the empirical usefulness of the proposed method.

We propose a semiparametric procedure for estimating high dimensional sparse inverse covariance matrix. Our method, named ALICE, is applicable to the elliptical family. Computationally, we develop an efficient dual inexact iterative projection (${\rm D_2}$P) algorithm based on the alternating direction method of multipliers (ADMM). Theoretically, we prove that the ALICE estimator achieves the parametric rate of convergence in both parameter estimation and model selection. Moreover, ALICE calibrates regularizations when estimating each column of the inverse covariance matrix. So it not only is asymptotically tuning free, but also achieves an improved finite sample performance. We present numerical simulations to support our theory, and a real data example to illustrate the effectiveness of the proposed estimator.

We propose two new principal component analysis methods in this paper utilizing a semiparametric model. The according methods are named Copula Component Analysis (COCA) and Copula PCA. The semiparametric model assumes that, af- ter unspecified marginally monotone transformations, the distributions are multi- variate Gaussian. The COCA and Copula PCA accordingly estimate the leading eigenvectors of the correlation and covariance matrices of the latent Gaussian dis- tribution. The robust nonparametric rank-based correlation coefficient estimator, Spearman’s rho, is exploited in estimation. We prove that, under suitable condi- tions, although the marginal distributions can be arbitrarily continuous, the COCA and Copula PCA estimators obtain fast estimation rates and are feature selection consistent in the setting where the dimension is nearly exponentially large relative to the sample size. Careful numerical experiments on the synthetic and real data are conducted to back up the theoretical results. We also discuss the relationship with the transelliptical component analysis proposed by Han and Liu (2012).

We propose a high dimensional semiparametric scale-invariant principle compo- nent analysis, named TCA, by utilize the natural connection between the ellipti- cal distribution family and the principal component analysis. Elliptical distribu- tion family includes many well-known multivariate distributions like multivari- ate Gaussian, t and logistic and it is extended to the meta-elliptical by Fang et.al (2002) using the copula techniques. In this paper we extend the meta-elliptical distribution family to a even larger family, called transelliptical. We prove that TCA can obtain a near-optimal splog d/n estimation consistency rate in recover- ing the leading eigenvector of the latent generalized correlation matrix under the transelliptical distribution family, even if the distributions are very heavy-tailed, have infinite second moments, do not have densities and possess arbitrarily con- tinuous marginal distributions. A feature selection result with explicit rate is also provided. TCA is further implemented in both numerical simulations and large- scale stock data to illustrate its empirical usefulness. Both theories and experi- ments confirm that TCA can achieve model flexibility, estimation accuracy and robustness at almost no cost.

We advocate the use of a new distribution family—the transelliptical—for robust inference of high dimensional graphical models. The transelliptical family is an extension of the nonparanormal family proposed by Liu et al. (2009). Just as the nonparanormal extends the normal by transforming the variables using univariate functions, the transelliptical extends the elliptical family in the same way. We propose a nonparametric rank-based regularization estimator which achieves the parametric rates of convergence for both graph recovery and parameter estima- tion. Such a result suggests that the extra robustness and flexibility obtained by the semiparametric transelliptical modeling incurs almost no efficiency loss. We also discuss the relationship between this work with the transelliptical component analysis proposed by Han and Liu (2012).

We prove a new exponential concentration inequality for a plug-in estimator of the Shannon mutual information. Previous results on mutual information estimation only bounded expected error. The advantage of having the exponential inequality is that, combined with the union bound, we can guarantee accurate estimators of the mutual information for many pairs of random variables simultaneously. As an application, we show how to use such a result to optimally estimate the density function and graph of a distribution which is Markov to a forest graph.

Many statistical methods gain robustness and exibility by sacricing convenient computational structure. In this paper, we illustrate this fundamental tradeoff by studying a semiparametric graphical model estimation problem. We explain how new computational techniques help to solve this type of problem. In particularly, we propose a smooth-projected neighborhood pursuit method for efciently estimating high dimensional nonparanormal graphs with theoretical guarantees. Besides new computational and theoretical analysis, we also provide an alternative view to analyze the tradeoff between computational efciency and statistical error under a smoothing optimization framework. We also report experimental results on text and stock datasets.

An important challenge in understanding climate change is to uncover the dependency relationships between various climate observations and forcing factors. Graphical lasso, a recently proposed L1 penalty based structure learning algorithm, has been proven successful for learning underlying dependency structures for the data drawn from a multivariate Gaussian distribution. However, climatological data often turn out to be non-Gaussian, e.g. cloud cover, precipitation, etc. In this paper, we examine nonparametric learning methods to address this challenge. In particular, we develop a methodology to learn dynamic graph structures from spatial-temporal data so that the graph structures at adjacent time or locations are similar. Experimental results demonstrate that our method not only recovers the underlying graph well but also captures the smooth variation properties on both synthetic data and climate data. An important challenge in understanding climate change is to uncover the dependency relationships between various climate observations and forcing factors. Graphical lasso, a recently proposed An important challenge in understanding climate change is to uncover the dependency relationships between various climate observations and forcing factors. Graphical lasso, a recently proposed L1 penalty based structure learning algorithm, has been proven successful for learning underlying dependency structures for the data drawn from a multivariate Gaussian distribution. However, climatological data often turn out to be non-Gaussian, e.g. cloud cover, precipitation, etc. In this paper, we examine nonparametric learning methods to address this challenge. In particular, we develop a methodology to learn dynamic graph structures from spatial-temporal data so that the graph structures at adjacent time or locations are similar. Experimental results demonstrate that our method not only recovers the underlying graph well but also captures the smooth variation properties on both synthetic data and climate data.

We propose a new nonparametric learning method based on multivariate dyadic regression trees (MDRTs). Unlike traditional dyadic decision trees (DDTs) or classification and regression trees (CARTs), MDRTs are constructed using penalized empirical risk minimization with a novel sparsity-inducing penalty. Theoretically, we show that MDRTs can simultaneously adapt to the unknown sparsity and smoothness of the true regression functions, and achieve the nearly optimal rates of convergence (in a minimax sense) for the class of $(\alpha, C)$-smooth functions. Empirically, MDRTs can simultaneously conduct function estimation and variable selection in high dimensions. To make MDRTs applicable for large-scale learning problems, we propose a greedy heuristics. The superior performance of MDRTs are demonstrated on both synthetic and real datasets.

Undirected graphical models encode in a graph $G$ the dependency structure of a random vector $Y$. In many applications, it is of interest to model $Y$ given another random vector $X$ as input. We refer to the problem of estimating the graph $G(x)$ of $Y$ conditioned on $X=x$ as ``graph-valued regression''. In this paper, we propose a semiparametric method for estimating $G(x)$ that builds a tree on the $X$ space just as in CART (classification and regression trees), but at each leaf of the tree estimates a graph. We call the method ``Graph-optimized CART'', or Go-CART. We study the theoretical properties of Go-CART using dyadic partitioning trees, establishing oracle inequalities on risk minimization and tree partition consistency. We also demonstrate the application of Go-CART to a meteorological dataset, showing how graph-valued regression can provide a useful tool for analyzing complex data.

A challenging problem in estimating high-dimensional graphical models is to choose the regularization parameter in a data-dependent way. The standard techniques include $K$-fold cross-validation ($K$-CV), Akaike information criterion (AIC), and Bayesian information criterion (BIC). Though these methods work well for low-dimensional problems, they are not suitable in high dimensional settings. In this paper, we present StARS: a new stability-based method for choosing the regularization parameter in high dimensional inference for undirected graphs. The method has a clear interpretation: we use the least amount of regularization that simultaneously makes a graph sparse and replicable under random sampling. This interpretation requires essentially no conditions. Under mild conditions, we show that StARS is partially sparsistent in terms of graph estimation: i.e. with high probability, all the true edges will be included in the selected model even when the graph size asymptotically increases with the sample size. Empirically, the performance of StARS is compared with the state-of-the-art model selection procedures, including $K$-CV, AIC, and BIC, on both synthetic data and a real microarray dataset. StARS outperforms all competing procedures.

Web search engines can perform poorly for long queries (i.e., those containing four or more terms), in part because of their high level of query specificity. The automatic assignment of labels to long queries can capture aspects of a user’s search intent that may not be apparent from the terms in the query. This affords search result matching or reranking based on queries and labels rather than the query text alone. Query labels can be derived from interaction logs generated from many users’ search result clicks or from query trails comprising the chain of URLs visited following query submission. However, since long queries are typically rare, they are difficult to label in this way because little or no historic log data exists for them. A subset of these queries may be amenable to labeling by detecting similarities between parts of a long and rare query and the queries which appear in logs. In this article, we present the comparison of four similarity algorithms for the automatic assignment of Open Directory Project category labels to long and rare queries, based solely on matching against similar satisfied query trails extracted from log data. Our findings show that although the similarity-matching algorithms we investigated have tradeoffs in terms of coverage and accuracy, one algorithm that bases similarity on a popular search result ranking function (effectively regarding potentially-similar queries as “documents”) outperforms the others. We find that it is possible to correctly predict the top label better than one in five times, even when no past query trail exactly matches the long and rare query. We show that these labels can be used to reorder top-ranked search results leading to a significant improvement in retrieval performance over baselines that do not utilize query labeling, but instead rank results using content-matching or click-through logs. The outcomes of our research have implications for search providers attempting to provide users with highly-relevant search results for long queries.

We develop a cyclical blockwise coordinate descent algorithm for the multi-task Lasso that efficiently solves problems with thousands of features and tasks. The main result shows that a closed-form Winsorization operator can be obtained for the sup-norm penalized least squares regression. This allows the algorithm to find solutions to very large-scale problems far more efficiently than existing methods. This result complements the pioneering work of Friedman, et al. (2007) for the single-task Lasso. As a case study, we use the multi-task Lasso as a variable selector to discover a semantic basis for predicting human neural activation. The learned solution outperforms the standard basis for this task on the majority of test participants, while requiring far fewer assumptions about cognitive neuroscience. We demonstrate how this learned basis can yield insights into how the brain represents the meanings of words.

This paper studies the forward greedy strategy in sparse nonparametric regression. For additive models, we propose an algorithm called additive forward regression; for general multivariate regression, we propose an algorithm called generalized forward regression. Both of them simultaneously conduct estimation and variable selection in nonparametric settings for the high dimensional sparse learning problem. Our main emphasis is empirical: on both simulated and real data, these two simple greedy methods can clearly outperform several state-of-the-art competitors, including the LASSO, a nonparametric version of the LASSO called the sparse additive model (SpAM) and a recently proposed adaptive parametric forward-backward algorithm called the Foba. Some theoretical justifications are also provided.

We propose new families of models and algorithms for high-dimensional nonparametric learning with joint sparsity constraints. Our approach is based on a regularization method that enforces common sparsity patterns across different function components in a nonparametric additive model. The algorithms employ a coordinate descent approach that is based on a functional soft-thresholding operator. The framework yields several new models, including multi-task sparse additive models, multi-response sparse additive models, and sparse additive multi-category logistic regression. The methods are illustrated with experiments on synthetic data and gene microarray data.

We present a new class of models for high-dimensional nonparametric regression and classification called sparse additive models (SpAM). Our methods combine ideas from sparse linear modeling and additive nonparametric regression. We de- rive a method for fitting the models that is effective even when the number of covariates is larger than the sample size. A statistical analysis of the properties of SpAM is given together with empirical results on synthetic and real data, show- ing that SpAM can be effective in fitting sparse nonparametric models in high dimensional data.

This paper addresses a central problem of Proteomics: estimating the amounts of each of the thousands of proteins in a cell culture or tissue sample. Although laboratory methods involving isotopes have been developed for this problem, we seek a method that uses simpler laboratory procedures. Specifically, our aim is to use data-mining techniques to infer protein levels from the relatively cheap and abundant data available from high-throughput tandem mass spectrometry (MS/MS). We have developed and evaluated several techniques for tackling this problem, including the development of three generative models of MS/MS data, and methods for efficiently fitting the models to data. In addition, we tested each method on three real-world datasets generated by MS/MS experiments performed on various tissue samples taken from Mouse. This paper outlines the biological problem and presents a selection of our results.

In this article, we propose a framework, called XAR-Miner, for mining ARs from XML documents efficiently. In XAR-Miner, raw data in the XML document first are preprocessed to transform either to an Indexed XML Tree (IX-tree) or to Multirelational Databases (Multi-DB), depending on the size of the XML document and the memory constraint of the system, for efficient data selection and AR mining. Concepts that are relevant to the AR mining task are generalized to produce generalized metapatterns. A suitable metric is devised for measuring the degree of concept generalization in order to prevent undergeneralization or overgeneralization. Resulting generalized metapatterns are used to generate large ARs that meet the support and confidence levels. A greedy algorithm is also presented in order to integrate data selection and large itemset generation to enhance the efficiency of the AR mining process. The experiments conducted show that XAR-Miner is more efficient in performing a large number of AR mining tasks from XML documents than the state-of-the-art method of repetitively scanning through XML documents in order to perform each of the mining tasks.