Jianwen Chen, Xinyu Yang, Peng Xia, Arian Azarang, Yueh Z Lee, Gang Li, Hongtu Zhu, Yun Li, Beidi Chen, Huaxiu Yao. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Relation extraction (RE) is a core task in natural language processing.Traditional approaches typically frame RE as a supervised learning problem, directly mapping context to labels-an approach that often suffers from poor out-of-domain (OOD) generalization.Inspired by the workflow of human annotators, we reframe RE as a reasoning task guided by annotation guidelines and introduce R1-RE, the first reinforcement learning with verifiable reward (RLVR) framework for RE tasks.Our method elicits the reasoning abilities of small language models for annotation tasks, resulting in significantly improved OOD robustness.We evaluate our approach on the public Sem-2010 dataset and a private MDKG dataset.The R1-RE-7B model attains an average OOD accuracy of approximately 70%, on par with leading proprietary models such as GPT-4o.Additionally, our comprehensive analysis provides novel insights into the training dynamics and emergent reasoning behaviors of the RLVR paradigm for RE. 1

In the field of machine learning, the pursuit of accurate models is ongoing. A key aspect of improving prediction performance lies in identifying which data points in the training set should be excluded and which high-quality, potentially unlabeled data points outside the training set should be incorporated to improve the model's performance on unseen data. To accomplish this, an effective metric is needed to evaluate the contribution of each data point toward enhancing overall model performance. This paper proposes the use of an influence measure as a metric to assess the impact of training data on test set performance. Additionally, we introduce a data selection method to optimize the training set as well as a dynamic active learning algorithm driven by the influence measure. The effectiveness of these methods is demonstrated through extensive simulations and real-world datasets.

A/B testing has become the gold standard for modern technological industries for policy evaluation. Motivated by the widespread use of switchback experiments in A/B testing, this paper conducts a comprehensive comparative analysis of various switchback designs in Markovian environments. Unlike many existing works which derive the optimal design based on specific and relatively simple estimators, our analysis covers a range of state-of-the-art estimators developed in the reinforcement learning literature. It reveals that the effectiveness of different switchback designs depends crucially on (i) the size of the carryover effect and (ii) the autocorrelations among reward errors over time. Meanwhile, these findings are estimator-agnostic, i.e., they apply to all the aforementioned estimators. Based on these insights, we provide a workflow to offer guidelines for practitioners on designing switchback experiments in A/B testing.

The advancement of Large Vision-Language Models (LVLMs) has propelled their application in the medical field. However, Medical LVLMs (Med-LVLMs) encounter factuality challenges due to modality misalignment, where the models prioritize textual knowledge over visual input, leading to hallucinations that contradict information in medical images. Previous attempts to enhance modality alignment in Med-LVLMs through preference optimization have inadequately addressed clinical relevance in preference data, making these samples easily distinguishable and reducing alignment effectiveness. In response, we propose MMedPO, a novel multimodal medical preference optimization approach that considers the clinical relevance of preference samples to enhance Med-LVLM alignment. MMedPO curates multimodal preference data by introducing two types of dispreference: (1) plausible hallucinations injected through target Med-LVLMs or GPT-4o to produce medically inaccurate responses, and (2) lesion region neglect achieved through local lesion-noising, disrupting visual understanding of critical areas. We then calculate clinical relevance for each sample based on scores from multiple Med-LLMs and visual tools, enabling effective alignment. Our experiments demonstrate that MMedPO significantly enhances factual accuracy in Med-LVLMs, achieving substantial improvements over existing preference optimization methods by 14.2% and 51.7% on the Med-VQA and report generation tasks, respectively. Our code are available in https://github.com/aiming-lab/MMedPO}{https://github.com/aiming-lab/MMedPO.

In recent years, two-sided marketplaces have emerged as viable business models in many real-world applications. In particular, we have moved from the social network paradigm to a network with two distinct types of participants representing the supply and demand of a specific good. Examples of industries include but are not limited to accommodation (Airbnb, Booking.com), video content (YouTube, Instagram, TikTok), ridesharing (Uber, Lyft), online shops (Etsy, Ebay, Facebook Marketplace), music (Spotify, Amazon), app stores (Apple App Store, Google App Store) or job sites (LinkedIn). The traditional research in most of these industries focused on satisfying the demand. OTAs would sell hotel accommodation, TV networks would broadcast their own content, or taxi companies would own their own vehicle fleet. In modern examples like Airbnb, YouTube, Instagram, or Uber, the platforms operate by outsourcing the service they provide to their users, whether they are hosts, content creators or drivers, and have to develop their models considering their needs and goals.

In this paper, we propose a novel framework, the Sampling-guided Heterogeneous Graph Neural Network (HT-GNN), to effectively tackle the challenge of missing data imputation in longitudinal studies. Unlike traditional methods, which often require extensive preprocessing to handle irregular or inconsistent missing data, our approach accommodates arbitrary missing data patterns while maintaining computational efficiency. HT-GNN models both observations and covariates as distinct node types, connecting observation nodes at successive time points through subject-specific longitudinal subnetworks, while covariate-observation interactions are represented by attributed edges within bipartite graphs. By leveraging subject-wise mini-batch sampling and a multi-layer temporal smoothing mechanism, HT-GNN efficiently scales to large datasets, while effectively learning node representations and imputing missing data. Extensive experiments on both synthetic and real-world datasets, including the Alzheimer's Disease Neuroimaging Initiative (DNI) dataset, demonstrate that HT-GNN significantly outperforms existing imputation methods, even with high missing data rates (e.g., 80%). The empirical results highlight HT-GNN's robust imputation capabilities and superior performance, particularly in the context of complex, large-scale longitudinal data.

Traditional functional linear regression usually takes a one-dimensional functional predictor as input and estimates the continuous coefficient function. Modern applications often generate two-dimensional covariates, which become matrices when observed at grid points. To avoid the inefficiency of the classical method involving estimation of a two-dimensional coefficient function, we propose a functional bilinear regression model, and introduce an innovative three-term penalty to impose roughness penalty in the estimation. The proposed estimator exhibits minimax optimal property for prediction under the framework of reproducing kernel Hilbert space. An iterative generalized cross-validation approach is developed to choose tuning parameters, which significantly improves the computational efficiency over the traditional cross-validation approach. The statistical and computational advantages of the proposed method over existing methods are further demonstrated via simulated experiments, the Canadian weather data, and a biochemical long-range infrared light detection and ranging data.

The recent emergence of Medical Large Vision Language Models (Med-LVLMs) has enhanced medical diagnosis.However, current Med-LVLMs frequently encounter factual issues, often generating responses that do not align with established medical facts.Retrieval-Augmented Generation (RAG), which utilizes external knowledge, can improve the factual accuracy of these models but introduces two major challenges.First, limited retrieved contexts might not cover all necessary information, while excessive retrieval can introduce irrelevant and inaccurate references, interfering with the model's generation.Second, in cases where the model originally responds correctly, applying RAG can lead to an over-reliance on retrieved contexts, resulting in incorrect answers.To address these issues, we propose RULE, which consists of two components.First, we introduce a provably effective strategy for controlling factuality risk through the calibrated selection of the number of retrieved contexts.Second, based on samples where over-reliance on retrieved contexts led to errors, we curate a preference dataset to fine-tune the model, balancing its dependence on inherent knowledge and retrieved contexts for generation.We demonstrate the effectiveness of RULE on medical VQA and report generation tasks across three datasets, achieving an average improvement of 47.4% in factual accuracy.We publicly release our benchmark and code in https: //github.com/richard-peng-xia/RULE.

This paper studies policy evaluation with multiple data sources, especially in scenarios that involve one experimental dataset with two arms, complemented by a historical dataset generated under a single control arm. We propose novel data integration methods that linearly integrate base policy value estimators constructed based on the experimental and historical data, with weights optimized to minimize the mean square error (MSE) of the resulting combined estimator. We further apply the pessimistic principle to obtain more robust estimators, and extend these developments to sequential decision making. Theoretically, we establish non-asymptotic error bounds for the MSEs of our proposed estimators, and derive their oracle, efficiency and robustness properties across a broad spectrum of reward shift scenarios. Numerical experiments and real-data-based analyses from a ridesharing company demonstrate the superior performance of the proposed estimators.

Artificial intelligence has significantly impacted medical applications, particularly with the advent of Medical Large Vision Language Models (Med-LVLMs), sparking optimism for the future of automated and personalized healthcare. However, the trustworthiness of Med-LVLMs remains unverified, posing significant risks for future model deployment. In this paper, we introduce CARES and aim to comprehensively evaluate the Trustworthiness of Med-LVLMs across the medical domain. We assess the trustworthiness of Med-LVLMs across five dimensions, including trustfulness, fairness, safety, privacy, and robustness. CARES comprises about 41K question-answer pairs in both closed and open-ended formats, covering 16 medical image modalities and 27 anatomical regions. Our analysis reveals that the models consistently exhibit concerns regarding trustworthiness, often displaying factual inaccuracies and failing to maintain fairness across different demographic groups. Furthermore, they are vulnerable to attacks and demonstrate a lack of privacy awareness. We publicly release our benchmark and code in https://github.com/richard-peng-xia/CARES.

MOTIVATION: Imaging genetics integrates imaging and genetic techniques to examine how genetic variations influence the function and structure of organs like the brain or heart, providing insights into their impact on behavior and disease phenotypes. The use of organ-wide imaging endophenotypes has increasingly been used to identify potential genes associated with complex disorders. However, analyzing organ-wide imaging data alongside genetic data presents two significant challenges: high dimensionality and complex relationships. To address these challenges, we propose a novel, nonlinear inference framework designed to partially mitigate these issues. RESULTS: We propose a functional partial least squares through distance covariance (FPLS-DC) framework for efficient genome wide analyses of imaging phenotypes. It consists of two components. The first component utilizes the FPLS-derived base functions to reduce image dimensionality while screening genetic markers. The second component maximizes the distance correlation between genetic markers and projected imaging data, which is a linear combination of the FPLS-basis functions, using simulated annealing algorithm. In addition, we proposed an iterative FPLS-DC method based on FPLS-DC framework, which effectively overcomes the influence of inter-gene correlation on inference analysis. We efficiently approximate the null distribution of test statistics using a gamma approximation. Compared to existing methods, FPLS-DC offers computational and statistical efficiency for handling large-scale imaging genetics. In real-world applications, our method successfully detected genetic variants associated with the hippocampus, demonstrating its value as a statistical toolbox for imaging genetic studies. AVAILABILITY AND IMPLEMENTATION: The FPLS-DC method we propose opens up new research avenues and offers valuable insights for analyzing functional and high-dimensional data. In addition, it serves as a useful tool for scientific analysis in practical applications within the field of imaging genetics research. The R package FPLS-DC is available in Github: https://github.com/BIG-S2/FPLSDC.

The conditional randomization test (CRT) was recently proposed to test whether two random variables X and Y are conditionally independent given random variables Z. The CRT assumes that the conditional distribution of X given Z is known under the null hypothesis and then it is compared to the distribution of the observed samples of the original data. The aim of this paper is to develop a novel alternative of CRT by using nearest-neighbor sampling without assuming the exact form of the distribution of X given Z. Specifically, we utilize the computationally efficient 1-nearest-neighbor to approximate the conditional distribution that encodes the null hypothesis. Then, theoretically, we show that the distribution of the generated samples is very close to the true conditional distribution in terms of total variation distance. Furthermore, we take the classifier-based conditional mutual information estimator as our test statistic. The test statistic as an empirical fundamental information theoretic quantity is able to well capture the conditional-dependence feature. We show that our proposed test is computationally very fast, while controlling type I and II errors quite well. Finally, we demonstrate the efficiency of our proposed test in both synthetic and real data analyses.

Reinforcement learning (RL) has made significant advancements in artificial intelligence. However, its real-world applications are limited due to differences between simulated environments and the actual world. Consequently, it is crucial to systematically analyze how each component of the RL system can affect the final model performance. In this study, we propose an adversarial learning framework for distributional reinforcement learning, which adopts the concept of influence measure from the statistics community. This framework enables us to detect performance loss caused by either the internal policy structure or the external state observation. The proposed influence measure is based on information geometry and has desirable properties of invariance. We demonstrate that the influence measure is useful for three diagnostic tasks: identifying fragile states in trajectories, determining the instability of the policy architecture, and pinpointing anomalously sensitive policy parameters.

Controlling the degree of stylization in the Neural Style Transfer (NST) is a little tricky since it usually needs hand-engineering on hyper-parameters. In this paper, we propose the first deep Reinforcement Learning (RL) based architecture that splits one-step style transfer into a step-wise process for the NST task. Our RL-based method tends to preserve more details and structures of the content image in early steps, and synthesize more style patterns in later steps. It is a user-easily-controlled style-transfer method. Additionally, as our RL-based model performs the stylization progressively, it is lightweight and has lower computational complexity than existing one-step Deep Learning (DL) based models. Experimental results demonstrate the effectiveness and robustness of our method.

Online marketplace is a digital platform that connects buyers (demand) and sellers (supply) and provides exposure opportunities that individual participants would not otherwise have access to. Online marketplaces exist in a diverse set of domains and industries, for example, rideshare (Lyft, DiDi, Uber), house rental (Airbnb), real estate (Beke), online retail (Amazon, Ebay), and food ordering and delivery (Doordash, Meituan). Besides academia, many companies and institutions are researching on topics specific to their particular domains. The fundamental mechanism of an online marketplace is to match supply and demand to generate transactions, with objectives considering service quality, participants experience, financial and operational efficiency. It is valuable to bring together researchers and practitioners from different application domains to discuss their experiences, challenges, and opportunities to leverage cross-domain knowledge. The goal of this workshop is to offer an opportunity to appreciate the diversity in applications, to draw connections to inform decision optimization across different industries, and to discover new problems that are fundamental to marketplaces of different domains. The previous version of this workshop at KDD-2022 was a tremendous success in terms of participation, technical contribution, and community interest. This updated version of the workshop is especially timely to cover the issues and algorithms pertinent to general online marketplaces, specific problems and applications arising from those diverse domains, as well as emerging topics such as competition and resilience to market condition shifts.

There is a growing interest in developing methods to estimate individualized treatment effects (ITEs) for various real-world applications, such as e-commerce and public health. This paper presents a novel architecture, called DNet, to infer distributional ITEs. DNet can learn the entire outcome distribution for each treatment, whereas most existing methods primarily focus on the conditional average treatment effect and ignore the conditional variance around its expectation. Additionally, our method excels in settings with heavy-tailed outcomes and outperforms state-of-the-art methods in extensive experiments on benchmark and real-world datasets. DNet has also been successfully deployed in a widely used mobile app with millions of daily active users.

A/B testing is critical for modern technological companies to evaluate the effectiveness of newly developed products against standard baselines. This paper studies optimal designs that aim to maximize the amount of information obtained from online experiments to estimate treatment effects accurately. We propose three optimal allocation strategies in a dynamic setting where treatments are sequentially assigned over time. These strategies are designed to minimize the variance of the treatment effect estimator when data follow a non Markov decision process or a (time-varying) Markov decision process. We further develop estimation procedures based on existing off-policy evaluation (OPE) methods and conduct extensive experiments in various environments to demonstrate the effectiveness of the proposed methodologies. In theory, we prove the optimality of the proposed treatment allocation design and establish upper bounds for the mean squared errors of the resulting treatment effect estimators.

Conditional independence (CI) testing is a fundamental task in statistics and machine learning, but its effectiveness is hindered by the challenges posed by high-dimensional conditioning variables and limited data samples. This article introduces a novel testing approach to address these challenges and enhance control of the type I error while achieving high power under alternative hypotheses. The proposed approach incorporates a computationally efficient classifier-based conditional mutual information (CMI) estimator, capable of capturing intricate dependence structures among variables. To approximate a distribution encoding the null hypothesis, a $k$-nearest-neighbor local sampling strategy is employed. An important advantage of this approach is its ability to operate without assumptions about distribution forms or feature dependencies. Furthermore, it eliminates the need to derive asymptotic null distributions for the estimated CMI and avoids dataset splitting, making it particularly suitable for small datasets. The method presented in this article demonstrates asymptotic control of the type I error and consistency against all alternative hypotheses. Extensive analyses using both synthetic and real data highlight the computational efficiency of the proposed test. Moreover, it outperforms existing state-of-the-art methods in terms of type I and II errors, even in scenarios with high-dimensional conditioning sets. Additionally, the proposed approach exhibits robustness in the presence of heavy-tailed data.

Deep neural networks (DNNs) have achieved great success in image classification, but can be very vulnerable to adversarial attacks with small perturbations to images. To improve adversarial image generation for DNNs, we develop a novel method, called mFI-PSO, which utilizes a Manifold-based First-order Influence measure for vulnerable image and pixel selection and the Particle Swarm Optimization for various objective functions. Our mFI-PSO can thus effectively design adversarial images with flexible, customized options on the number of perturbed pixels, the misclassification probability, and the targeted incorrect class. Experiments demonstrate the flexibility and effectiveness of our mFI-PSO in adversarial attacks and its appealing advantages over some popular methods.

Online marketplace is a digital platform that connects buyers (demand) and sellers (supply) and provides exposure opportunities that individual participants would not otherwise have access to. Online marketplaces exist in a diverse set of domains and industries, for example, rideshare (Lyft, DiDi, Uber), house rental (Airbnb), real estate (Beke), online retail (Amazon, Ebay), job search (LinkedIn, Indeed.com, CareerBuilder), and food ordering and delivery (Doordash, Meituan). Besides academia, many companies and institutions are researching on topics specific to their particular domains. The fundamental mechanism of an online marketplace is to match supply and demand to generate transactions, with objectives considering service quality, participants experience, financial and operational efficiency. It is valuable to bring together researchers and practitioners from different application domains to discuss their experiences, challenges, and opportunities to leverage cross-domain knowledge. The goal of this workshop is to offer an opportunity to appreciate the diversity in applications, to draw connections to inform decision optimization across different industries, and to discover new problems that are fundamental to marketplaces of different domains. This workshop will follow a dual-track format. Track 1 covers the issues and algorithms pertinent to general online marketplaces as well as specific problems and applications arising from those diverse domains, such as ridesharing, online retail, food delivery, house rental, real estate, and more. Track 2 focuses on the state of the art advances in the computational jobs marketplace. Interesting challenges in this domain include the drastic increase of work from home or remote work, the imbalance between the demand and supply of the job market, the popularity of independent workers, the capability of helping job seekers on their whole job seeking journey and career development, the different objectives and behaviors of all major stakeholders in the ecosystem, e.g. job seekers, employers, recruiters and job agents.

Spatial transcriptomics (ST) technologies allow researchers to examine transcriptional profiles along with maintained positional information. Such spatially resolved transcriptional characterization of intact tissue samples provides an integrated view of gene expression in its natural spatial and functional context. However, high-throughput sequencing-based ST technologies cannot yet reach single cell resolution. Thus, similar to bulk RNA-seq data, gene expression data at ST spot-level reflect transcriptional profiles of multiple cells and entail the inference of cell-type composition within each ST spot for valid and powerful subsequent analyses. Realizing the critical importance of cell-type decomposition, multiple groups have developed ST deconvolution methods. The aim of this work is to review state-of-the-art methods for ST deconvolution, comparing their strengths and weaknesses. In particular, we construct ST spots from single-cell level ST data to assess the performance of 10 methods, with either ideal reference or non-ideal reference. Furthermore, we examine the performance of these methods on spot- and bead-level ST data by comparing estimated cell-type proportions to carefully matched single-cell ST data. In comparing the performance on various tissues and technological platforms, we concluded that RCTD and stereoscope achieve more robust and accurate inferences.

Modern biomedical studies often collect multi-view data, that is, multiple types of data measured on the same set of objects. A popular model in high-dimensional multi-view data analysis is to decompose each view’s data matrix into a low-rank common-source matrix generated by latent factors common across all data views, a low-rank distinctive-source matrix corresponding to each view, and an additive noise matrix. We propose a novel decomposition method for this model, called decomposition-based generalized canonical correlation analysis (D-GCCA). The D-GCCA rigorously defines the decomposition on the L2 space of random variables in contrast to the Euclidean dot product space used by most existing methods, thereby being able to provide the estimation consistency for the low-rank matrix recovery. Moreover, to well calibrate common latent factors, we impose a desirable orthogonality constraint on distinctive latent factors. Existing methods, however, inadequately consider such orthogonality and may thus suffer from substantial loss of undetected common-source variation. Our D-GCCA takes one step further than generalized canonical correlation analysis by separating common and distinctive components among canonical variables, while enjoying an appealing interpretation from the perspective of principal component analysis. Furthermore, we propose to use the variable-level proportion of signal variance explained by common or distinctive latent factors for selecting the variables most influenced. Consistent estimators of our D-GCCA method are established with good finite-sample numerical performance, and have closed-form expressions leading to efficient computation especially for large-scale data. The superiority of D-GCCA over state-of-the-art methods is also corroborated in simulations and real-world data examples.

In a bike-sharing system, demand loss is primarily due to out-of-stock stations. One solution to tackle this problem is to rebalance the bike inventory of the stations through repositioning. In a large-scale bike-sharing system, bike repositioning typically works in two steps: The platform generates the reposition tasks, and then the operators execute those tasks. In this paper, we focus on the problem of determining the reposition tasks, which is a core problem for the platform operations. We model the bike-sharing system as a networked inventory system and propose a select-and-match method to generate optimal repositions, seamlessly combining inventory management and spatiotemporal value learning techniques. We compute the optimal (s, S) policy to select the stations with excess and deficit bike stocks. Subsequently, we solve a minimum cost maximum flow problem for rebalance assignment while considering the long-term effects of repositioning, characterized by a bike transition value. We design a spatiotemporal value network to learn the values. We evaluate the performance of the proposed method in a simulator based on real bike-sharing data, and the results demonstrate the superiority of our approach over the other alternatives.

The aim of this paper is to develop a multi-objective distributional reinforcement learning framework for improving order dispatching on large-scale ride-hailing platforms. Compared with traditional RL-based approaches that focus on drivers’ income, the proposed framework also accounts for the spatiotemporal difference between the supply and demand networks. Specifically, we model the dispatching problem as a two-objective Semi-Markov Decision Process (SMDP) and estimate the relative importance of the two objectives under some unknown existing policy via Inverse Reinforcement Learning (IRL). Then, we combine Implicit Quantile Networks (IQN) with the traditional Deep Q-Networks (DQN) to jointly learn the two return distributions and adjusting their weights to refine the old policy through on-line planning and achieve a higher supply-demand coherence of the platform. We conduct large-scale dispatching experiments to demonstrate the remarkable improvement of proposed approach on the platform’s efficiency.

Large ride-hailing platforms, such as DiDi, Uber and Lyft, connect tens of thousands of vehicles in a city to millions of ride demands throughout the day, providing great promises for improving transportation efficiency through the tasks of order dispatching and vehicle repositioning. Existing studies, however, usually consider the two tasks in simplified settings that hardly address the complex interactions between the two, the real-time fluctuations between supply and demand, and the necessary coordinations due to the large-scale nature of the problem. In this paper we propose a unified value-based dynamic learning framework (V1D3) for tackling both tasks. At the center of the framework is a globally shared value function that is updated continuously using online experiences generated from real-time platform transactions. To improve the sample-efficiency and the robustness, we further propose a novel periodic ensemble method combining the fast online learning with a large-scale offline training scheme that leverages the abundant historical driver trajectory data. This allows the proposed framework to adapt quickly to the highly dynamic environment, to generalize robustly to recurrent patterns and to drive implicit coordinations among the population of managed vehicles. Extensive experiments based on real-world datasets show considerably improvements over other recently proposed methods on both tasks. Particularly, V1D3 outperforms the first prize winners of both dispatching and repositioning tracks in the KDD Cup 2020 RL competition, achieving state-of-the-art results on improving both total driver income and user experience related metrics.

Large-scale matrix linear regression models with high-dimensional responses and high-dimensional variables have been widely employed in various large-scale biomedical studies. In this article, we propose an optimal minimax variable selection approach for the matrix linear regression model when the dimensions of both the response matrix and predictors diverge at the exponential rate of the sample size. We develop an iterative hard-thresholding algorithm for fast computation and establish an optimal minimax theory for the parameter estimates. The finite sample performance of the method is examined via extensive simulation studies and a real data application from the Alzheimer's Disease Neuroimaging Initiative study is provided.

Tech companies (e.g., Google or Facebook) often use randomized online experiments and/or A/B testing primarily based on the average treatment effects to compare their new product with an old one. However, it is also critically important to detect qualitative treatment effects such that the new one may significantly outperform the existing one only under some specific circumstances. The aim of this paper is to develop a powerful testing procedure to efficiently detect such qualitative treatment effects. We propose a scalable online updating algorithm to implement our test procedure. It has three novelties including adaptive randomization, sequential monitoring, and online updating with guaranteed type-I error control. We also thoroughly examine the theoretical properties of our testing procedure including the limiting distribution of test statistics and the justification of an efficient bootstrap method. Extensive empirical studies are conducted to examine the finite sample performance of our test procedure.

Users' behavioral predictions are crucially important for many domains including major e-commerce companies, ride-hailing platforms, social networking, and education. The success of such prediction strongly depends on the development of representation learning that can effectively model the dynamic evolution of user's behavior. This paper aims to develop a joint framework of combining inverse reinforcement learning (IRL) with deep learning (DL) regression model, called IRL-DL, to predict drivers' future behavior in ride-hailing platforms. Specifically, we formulate the dynamic evolution of each driver as a sequential decision-making problem and then employ IRL as representation learning to learn the preference vector of each driver. Then, we integrate drivers' preference vector with their static features (e.g., age, gender) and other attributes to build a regression model (e.g., LTSM-neural network) to predict drivers' future behavior. We use an extensive driver data set obtained from a ride-sharing platform to verify the effectiveness and efficiency of our IRL-DL framework, and results show that our IRL-DL framework can achieve consistent and remarkable improvements over models without drivers' preference vectors.

Deep neural networks (DNNs) have achieved superior performance in various prediction tasks, but can be very vulnerable to adversarial examples or perturbations. Therefore, it is crucial to measure the sensitivity of DNNs to various forms of perturbations in real applications. We introduce a novel perturbation manifold and its associated influence measure to quantify the effects of various perturbations on DNN classifiers. Such perturbations include various external and internal perturbations to input samples and network parameters. The proposed measure is motivated by information geometry and provides desirable invariance properties. We demonstrate that our influence measure is useful for four model building tasks: detecting potential ‘outliers’, analyzing the sensitivity of model architectures, comparing network sensitivity between training and test sets, and locating vulnerable areas. Experiments show reasonably good performance of the proposed measure for the popular DNN models ResNet50 and DenseNet121 on CIFAR10 and MNIST datasets.

We consider the problem of performing an association test between functional data and scalar variables in a varying coefficient model setting. We propose a functional projection regression model and an associated global test statistic to aggregate relatively weak signals across the domain of functional data, while reducing the dimension. An optimal functional projection direction is selected to maximize signal-to-noise ratio with ridge penalty. Theoretically, we systematically study the asymptotic distribution of the global test statistic and provide a strategy to adaptively select the optimal tuning parameter. We use simulations to show that the proposed test outperforms all existing state-of-the-art methods in functional statistical inference. Finally, we apply the proposed testing method to the genome-wide association analysis of imaging genetic data in UK Biobank dataset.

Origin-Destination (OD) flow data is an important instrument for traffic study and management. So far traditional ways like surveys or detectors are costly and only give limited availability of OD flows. Various statistical and stochastic models for OD flow estimation and prediction based on limited link volume data or automatic vehicle identification (AVI) data have been developed. However, smartphone-generated trajectory data has not been as much leveraged in this field, though the usage of smartphones in traveling is emerging in recent years. In this paper, we propose a semi-supervised deep learning based model that appropriately combines both AVI and smartphone trajectory data during training and is able to generate predictions of OD flows in an urban network solely based on the smartphone trajectory data at inference time. Our model can provide OD estimation and prediction services on larger spatial areas beyond the limited spatial coverage of AVI data. Tests of our model using real data have shown promising results, compared with an AVI input-dependent Kalman filter model. Potentially, our model can easily be embedded to a trajectory collecting platform and generate continuous real-time OD flow predictions online.

Data-enabled smart transportation has attracted a surge of interest from machine learning and data mining researchers nowadays due to the bloom of online ride-hailing industry and rapid development of autonomous driving. Large-scale high quality route data and trading data (spatiotemporal data) have been generated every day, which makes AI an urgent need and preferred solution for the decision making in intelligent transportation systems. While a large of amount of work have been dedicated to traditional transportation problems, they are far from satisfactory for the rising need. We propose a half-day workshop at CIKM 2019 for the professionals, researchers, and practitioners who are interested in mining and understanding big and heterogeneous data generated in transportation, and AI applications to improve the transportation system. We plan to have several invited talks from both academia and industry. This workshop would be organized by Shanghai Jiao Tong University, Didi Chuxing and Pennsylvania State University.

For both the traditional street-hailing taxi industry and the recently emerged on-line ride-hailing, it has been a major challenge to improve the ride-hailing marketplace efficiency due to spatio-temporal imbalance between the supply and demand, among other factors. Despite the numerous approaches to improve marketplace efficiency using pricing and dispatch strategies, they usually optimize pricing or dispatch separately. In this paper, we show that these two processes are in fact intrinsically interrelated. Motivated by this observation, we make an attempt to simultaneously optimize pricing and dispatch strategies. However, such a joint optimization is extremely challenging due to the inherent huge scale and lack of a uniform model of the problem. To handle the high complexity brought by the new problem, we propose InBEDE (Integrating contextual Bandit with tEmporal DiffErence learning), a learning framework where pricing strategies are learned via a contextual bandit algorithm, and the dispatch strategies are optimized with the help of temporal difference learning. The two learning components proceed in a mutual bootstrapping manner, in the sense that the policy evaluations of the two components are inter-dependent. Evaluated with real-world datasets of two Chinese cities from Didi Chuxing, an online ride-hailing platform, we show that the market efficiency of the ride-hailing platform can be significantly improved using InBEDE.

Recent works on ride-sharing order dispatching have highlighted the importance of taking into account both the spatial and temporal dynamics in the dispatching process for improving the transportation system efficiency. At the same time, deep reinforcement learning has advanced to the point where it achieves superhuman performance in a number of fields. In this work, we propose a deep reinforcement learning based solution for order dispatching and we conduct large scale online A/B tests on DiDi's ride-dispatching platform to show that the proposed method achieves significant improvement on both total driver income and user experience related metrics.

Graph-based semi-supervised learning is a very powerful tool in classification tasks, while in most existing literature the labelled nodes are assumed to be randomly sampled. When the labelling status depends on the unobserved node response, ignoring the missingness can lead to significant estimation bias and handicap the classifiers. This situation is called non-ignorable non-response. To solve the problem, we propose a Graph-based joint model with Non-ignorable Non-response (GNN), followed by a joint inverse weighting estimation procedure incorporated with sampling imputation approach. Our method is proved to outperform some state-of-art models in both regression and classification problems, by simulations and real analysis on the Cora dataset.

Ride dispatching is a central operation task on a ride-sharing platform to continuously match drivers to trip-requesting passengers. In this work, we model the ride dispatching problem as a Markov Decision Process and propose learning solutions based on deep Q-networks with action search to optimize the dispatching policy for drivers on ride-sharing platforms. We train and evaluate dispatching agents for this challenging decision task using real-world spatio-temporal trip data from the DiDi ride-sharing platform. A large-scale dispatching system typically supports many geographical locations with diverse demand-supply settings. To increase learning adaptability and efficiency, we propose a new transfer learning method Correlated Feature Progressive Transfer, along with two existing methods, enabling knowledge transfer in both spatial and temporal spaces. Through an extensive set of experiments, we demonstrate the learning and optimization capabilities of our deep reinforcement learning algorithms. We further show that dispatching policies learned by transferring knowledge from a source city to target cities or across temporal space within the same city significantly outperform those without transfer learning.

Motivation: The aim of precision medicine is to harness new knowledge and technology to optimize the timing and targeting of interventions for maximal therapeutic benefit. This study explores the possibility of building AI models without precise pixel-level annotation in prediction of the tumor size, extrathyroidal extension, lymph node metastasis, cancer stage and BRAF mutation in thyroid cancer diagnosis, providing the patients' background information, histopathological and immunohistochemical tissue images. Results: A novel framework for objective evaluation of automatic patient diagnosis algorithms has been established under the auspices of the IEEE International Symposium on Biomedical Imaging 2017- A Grand Challenge for Tissue Microarray Analysis in Thyroid Cancer Diagnosis. Here, we present the datasets, methods and results of the challenge and lay down the principles for future uses of this benchmark. The main contributions of the challenge include the creation of the data repository of tissue microarrays; the creation of the clinical diagnosis classification data repository of thyroid cancer; and the definition of objective quantitative evaluation for comparison and ranking of the algorithms. With this benchmark, three automatic methods for predictions of the five clinical outcomes have been compared, and detailed quantitative evaluation results are presented in this paper. Based on the quantitative evaluation results, we believe automatic patient diagnosis is still a challenging and unsolved problem. Availability and implementation: The datasets and the evaluation software will be made available to the research community, further encouraging future developments in this field. (http://www-o.ntust.edu.tw/cvmi/ISBI2017/). Contact: [email protected]. Supplementary information: Supplementary data are available at Bioinformatics online.

Magnetic resonance imaging (MRI) has become an important imaging technique for quantifying the spatial location and magnitude/direction of longitudinal cartilage morphology changes in patients with osteoarthritis (OA). Although several analytical methods, such as subregion-based analysis, have been developed to refine and improve quantitative cartilage analyses, they can be suboptimal due to two major issues: the lack of spatial correspondence across subjects and time and the spatial heterogeneity of cartilage progression across subjects. The aim of this paper is to present a statistical method for longitudinal cartilage quantification in OA patients, while addressing these two issues. The 3D knee image data is preprocessed to establish spatial correspondence across subjects and/or time. Then, a Gaussian hidden Markov model (GHMM) is proposed to deal with the spatial heterogeneity of cartilage progression across both time and OA subjects. To estimate unknown parameters in GHMM, we employ a pseudo-likelihood function and optimize it by using an expectation-maximization (EM) algorithm. The proposed model can effectively detect diseased regions in each OA subject and present a localized analysis of longitudinal cartilage thickness within each latent subpopulation. Our GHMM integrates the strengths of two standard statistical methods including the local subregion-based analysis and the ordered value approach. We use simulation studies and the Pfizer longitudinal knee MRI dataset to evaluate the finite sample performance of GHMM in the quantification of longitudinal cartilage morphology changes. Our results indicate that GHMM significantly outperforms several standard analytical methods.

Twin imaging studies have been valuable for understanding the relative contribution of the environment and genes on brain structures and their functions. Conventional analyses of twin imaging data include three sequential steps: spatially smoothing imaging data, independently fitting a structural equation model at each voxel, and finally correcting for multiple comparisons. However, conventional analyses are limited due to the same amount of smoothing throughout the whole image, the arbitrary choice of smoothing extent, and the decreased power in detecting environmental and genetic effects introduced by smoothing raw images. The goal of this paper is to develop a two-stage multiscale adaptive regression method (TwinMARM) for spatial and adaptive analysis of twin neuroimaging and behavioral data. The first stage is to establish the relationship between twin imaging data and a set of covariates of interest, such as age and gender. The second stage is to disentangle the environmental and genetic influences on brain structures and their functions. In each stage, TwinMARM employs hierarchically nested spheres with increasing radii at each location and then captures spatial dependence among imaging observations via consecutively connected spheres across all voxels. Simulation studies show that our TwinMARM significantly outperforms conventional analyses of twin imaging data. Finally, we use our method to detect statistically significant effects of genetic and environmental variations on white matter structures in a neonatal twin study.

We propose a novel diffusion tensor imaging (DTI) registration algorithm, called fast tensor image morphing for elastic registration (F-TIMER). F-TIMER leverages multiscale tensor regional distributions and local boundaries for hierarchically driving deformable matching of tensor image volumes. Registration is achieved by utilizing a set of automatically determined structural landmarks, via solving a soft correspondence problem. Based on the estimated correspondences, thin-plate splines are employed to generate a smooth, topology preserving, and dense transformation, and to avoid arbitrary mapping of nonlandmark voxels. To mitigate the problem of local minima, which is common in the estimation of high dimensional transformations, we employ a hierarchical strategy where a small subset of voxels with more distinctive attribute vectors are first deployed as landmarks to estimate a relatively robust low-degrees-of-freedom transformation. As the registration progresses, an increasing number of voxels are permitted to participate in refining the correspondence matching. A scheme as such allows less conservative progression of the correspondence matching towards the optimal solution, and hence results in a faster matching speed. Compared with its predecessor TIMER, which has been shown to outperform state-of-the-art algorithms, experimental results indicate that F-TIMER is capable of achieving comparable accuracy at only a fraction of the computation cost.

Diffusion tensor imaging (DTI) provides important information on the structure of white matter fiber bundles as well as detailed tissue properties along these fiber bundles in vivo. This paper presents a functional regression framework, called FRATS, for the analysis of multiple diffusion properties along fiber bundle as functions in an infinite dimensional space and their association with a set of covariates of interest, such as age, diagnostic status and gender, in real applications. The functional regression framework consists of four integrated components: the local polynomial kernel method for smoothing multiple diffusion properties along individual fiber bundles, a functional linear model for characterizing the association between fiber bundle diffusion properties and a set of covariates, a global test statistic for testing hypotheses of interest, and a resampling method for approximating the p-value of the global test statistic. The proposed methodology is applied to characterizing the development of five diffusion properties including fractional anisotropy, mean diffusivity, and the three eigenvalues of diffusion tensor along the splenium of the corpus callosum tract and the right internal capsule tract in a clinical study of neurodevelopment. Significant age and gestational age effects on the five diffusion properties were found in both tracts. The resulting analysis pipeline can be used for understanding normal brain development, the neural bases of neuropsychiatric disorders, and the joint effects of environmental and genetic factors on white matter fiber bundles.

Diffusion tensors are estimated from magnetic resonance images (MRIs) that are diffusion-weighted, and those images inherently contain noise. Therefore, noise in the diffusion-weighted images produces uncertainty in estimation of the tensors and their derived parameters, which include eigenvalues, eigenvectors, and the trajectories of fiber pathways that are reconstructed from those eigenvalues and eigenvectors. Although repetition and wild bootstrap methods have been widely used to quantify the uncertainty of diffusion tensors and their derived parameters, we currently lack theoretical derivations that would validate the use of these two bootstrap methods for the estimation of statistical parameters of tensors in the presence of noise. The aim of this paper is to examine theoretically and numerically the repetition and wild bootstrap methods for approximating uncertainty in estimation of diffusion tensor parameters under two different schemes for acquiring diffusion weighted images. Whether these bootstrap methods can be used to quantify uncertainty in some diffusion tensor parameters, such as fractional anisotropy (FA), depends critically on the morphology of the diffusion tensor that is being estimated. The wild and repetition bootstrap methods in particular cannot quantify uncertainty in the principal direction (PD) of isotropic (or oblate) tensor. We also examine the use of bootstrap methods in estimating tensors in a voxel containing multiple tensors, demonstrating their limitations when quantifying the uncertainty of tensor parameters in those locations. Simulation studies are also used to understand more thoroughly our theoretical results. Our findings raise serious concerns about the use of bootstrap methods to quantify the uncertainty of fiber pathways when those pathways pass through voxels that contain either isotropic tensors, oblate tensors, or multiple tensors.

Interest in the morphometric analysis of the brain and its subregions has recently intensified because growth or degeneration of the brain in health or illness affects not only the volume but also the shape of cortical and subcortical brain regions, and new image processing techniques permit detection of small and highly localized perturbations in shape or localized volume, with remarkable precision. An appropriate statistical representation of the shape of a brain region is essential, however, for detecting, localizing, and interpreting variability in its surface contour and for identifying differences in volume of the underlying tissue that produce that variability across individuals and groups of individuals. Our statistical representation of the shape of a brain region is defined by a reference region for that region and by a Gaussian random field (GRF) that is defined across the entire surface of the region. We first select a reference region from a set of segmented brain images of healthy individuals. The GRF is then estimated as the signed Euclidean distances between points on the surface of the reference region and the corresponding points on the corresponding region in images of brains that have been coregistered to the reference. Correspondences between points on these surfaces are defined through deformations of each region of a brain into the coordinate space of the reference region using the principles of fluid dynamics. The warped, coregistered region of each subject is then unwarped into its native space, simultaneously bringing into that space the map of corresponding points that was established when the surfaces of the subject and reference regions were tightly coregistered. The proposed statistical description of the shape of surface contours makes no assumptions, other than smoothness, about the shape of the region or its GRF. The description also allows for the detection and localization of statistically significant differences in the shapes of the surfaces across groups of subjects at both a fine and coarse scale. We demonstrate the effectiveness of these statistical methods by applying them to study differences in shape of the amygdala and hippocampus in a large sample of normal subjects and in subjects with attention deficit/hyperactivity disorder (ADHD).

Methods for the analysis of brain morphology, including voxel-based morphology and surface-based morphometries, have been used to detect associations between brain structure and covariates of interest, such as diagnosis, severity of disease, age, IQ, and genotype. The statistical analysis of morphometric measures usually involves two statistical procedures: 1) invoking a statistical model at each voxel (or point) on the surface of the brain or brain subregion, followed by mapping test statistics (e.g., t test) or their associated p values at each of those voxels; 2) correction for the multiple statistical tests conducted across all voxels on the surface of the brain region under investigation. We propose the use of new statistical methods for each of these procedures. We first use a heteroscedastic linear model to test the associations between the morphological measures at each voxel on the surface of the specified subregion (e.g., cortical or subcortical surfaces) and the covariates of interest. Moreover, we develop a robust test procedure that is based on a resampling method, called wild bootstrapping. This procedure assesses the statistical significance of the associations between a measure of given brain structure and the covariates of interest. The value of this robust test procedure lies in its computationally simplicity and in its applicability to a wide range of imaging data, including data from both anatomical and functional magnetic resonance imaging (fMRI). Simulation studies demonstrate that this robust test procedure can accurately control the family-wise error rate. We demonstrate the application of this robust test procedure to the detection of statistically significant differences in the morphology of the hippocampus over time across gender groups in a large sample of healthy subjects.