MOTIVATION: This work presents the first large-scale neutral benchmark experiment focused on single-event, right-censored, low-dimensional survival data. Benchmark experiments are essential in methodological research to scientifically compare new and existing model classes through proper empirical evaluation. Existing benchmarks in the survival literature are smaller in scale regarding the number of used datasets and extent of empirical evaluation. They often lack appropriate tuning or evaluation procedures, while other comparison studies focus on qualitative reviews rather than quantitative comparisons. This comprehensive study aims to fill the gap by neutrally evaluating a broad range of methods and providing generalizable guidelines for practitioners. RESULTS: We benchmark 21 models, ranging from classical statistical approaches to many common machine learning methods, on 34 publicly available datasets. The benchmark tunes models using both a discrimination measure (Harrell's C-index) and a scoring rule (Integrated Survival Brier Score), and evaluates them across six metrics covering discrimination, calibration, and overall predictive performance. Despite superior average ranks in overall predictive performance from individual learners like oblique random survival forests and likelihood-based boosting, and better discrimination rankings from multiple boosting- and tree-based methods as well as parametric survival models, no method statistically significantly outperforms the commonly used Cox proportional hazards model for either tuning measure. We conclude that while the Cox Proportional Hazards model remains a robust default for low-dimensional, right-censored survival data, more flexible methods may be preferable for specific dataset characteristics. AVAILABILITY AND IMPLEMENTATION: All code, data, and results are publicly available on GitHub https://github.com/slds-lmu/paper_2023_survival_benchmark and archived on Zenodo https://doi.org/10.5281/zenodo.19075310.

Abstract In recent years, overparametrization has received considerable attention in various fields, shown to accelerate training and promote simplicity. However, few works study the induced sparse regularization of the original parameters that is caused by combining overparametrization with explicit smooth regularization. Here, we present a unifying framework for smooth optimization of explicitly regularized objectives for (structured) sparsity. These non-smooth and possibly non-convex problems typically rely on solvers tailored to specific models or regularizers and have not been widely adopted in deep learning. In contrast, our method promises fully differentiable and approximation-free optimization for sparse regularizers and is thus compatible with the ubiquitous gradient descent paradigm. The proposed optimization transfer comprises overparameterization of selected parameters and a change of penalties. We prove that the surrogate objective is equivalent in the sense of identical global and local minima, thereby avoiding the introduction of spurious solutions. We comprehensively review sparsity-inducing parametrizations across different fields and combine them in our explicit surrogate regularization framework. We further extend their scope, point out improvements and present novel parametrizations. Numerical experiments further demonstrate the correctness and effectiveness of our approach on several sparse learning problems from high-dimensional regression to sparse neural network training.

Multi-epitope vaccines (EVs) represent a versatile and promising approach for combating a wide range of diseases, including viral and bacterial infections, parasitic diseases, and cancer. A critical aspect of EV design is ensuring efficient proteasomal cleavage of the synthetic polypeptide to recover therapeutic epitopes for presentation to$\mathbf{T}$cells. This process is influenced by the selection and arrangement of epitopes, as well as the design of linkers joining them. Modern EV design frameworks leverage proteasomal cleavage predictors to optimize epitope recovery and vaccine efficacy. However, the predictive power of these tools remains a limiting factor, particularly for challenging cleavage sites such as N -terminals. In this work, we systematically review and benchmark recent advances in proteasomal cleavage prediction, focusing on deep learning-based methods. We evaluate a range of architectures, including MLPs, CNNs, LSTMs, and Transformers, alongside simpler baselines such as logistic regression. Our results demonstrate that while complex models achieve marginally higher predictive performance, simpler models remain competitive and offer significant computational efficiency. We also explore the impact of dataset size, window size, and training techniques, finding diminishing returns from increasingly larger datasets and more complex models. Notably, our benchmarking highlights substantial improvements in predicting N-terminal cleavage sites, which are often overlooked but critical for EV design. Our findings provide practical guidance for the development of nextgeneration proteasomal cleavage predictors and underscore the importance of considering the probabilistic nature of cleavage in EV design. By consolidating the state of the art and introducing efficient baselines, this work aims to advance the field of multiepitope vaccine design and support the development of more effective and personalized immunotherapies.11Code for all experiments is available at this repository: https://github.com/ziegleringo/cleavage_benchmark.

Ensuring fairness in machine learning remains a significant challenge, as models often inherit biases from their training data. Generative models have recently emerged as a promising approach to mitigate bias at the data level while preserving utility. However, many rely on deep architectures, despite evidence that simpler models can be highly effective for tabular data. In this work, we introduce TabFairGDT, a novel method for generating fair synthetic tabular data using autoregressive decision trees. To enforce fairness, we propose a soft leaf resampling technique that adjusts decision tree outputs to reduce bias while preserving predictive performance. Our approach is non-parametric, effectively capturing complex relationships between mixed feature types, without relying on assumptions about the underlying data distributions. We evaluate TabFairGDT on benchmark fairness datasets and demonstrate that it outperforms state-of-the-art (SOTA) deep generative models, achieving better fairness-utility trade-off for downstream tasks, as well as higher synthetic data quality. Moreover, our method is lightweight, highly efficient, and CPU-compatible, requiring no data preprocessing. Remarkably, TabFairGDT achieves a 72% average speedup over the fastest SOTA baseline across various dataset sizes, and can generate fair synthetic data for medium-sized datasets (10 features, 10K samples) in just one second on a standard CPU, making it an ideal solution for realworld fairness-sensitive applications.

We discover a theoretical connection between explanation estimation and distribution compression that significantly improves the approximation of feature attributions, importance, and effects. While the exact computation of various machine learning explanations requires numerous model inferences and becomes impractical, the computational cost of approximation increases with an ever-increasing size of data and model parameters. We show that the standard i.i.d. sampling used in a broad spectrum of algorithms for post-hoc explanation leads to an approximation error worthy of improvement. To this end, we introduce Compress Then Explain (CTE), a new paradigm of sample-efficient explainability. It relies on distribution compression through kernel thinning to obtain a data sample that best approximates its marginal distribution. CTE significantly improves the accuracy and stability of explanation estimation with negligible computational overhead. It often achieves an on-par explanation approximation error 2-3x faster by using fewer samples, i.e. requiring 2-3x fewer model evaluations. CTE is a simple, yet powerful, plug-in for any explanation method that now relies on i.i.d. sampling.

Sparse regularization techniques are well-established in machine learning, yet their application in neural networks remains challenging due to the non-differentiability of penalties like the $L_1$ norm, which is incompatible with stochastic gradient descent. A promising alternative is shallow weight factorization, where weights are decomposed into two factors, allowing for smooth optimization of $L_1$-penalized neural networks by adding differentiable $L_2$ regularization to the factors. In this work, we introduce deep weight factorization, extending previous shallow approaches to more than two factors. We theoretically establish equivalence of our deep factorization with non-convex sparse regularization and analyze its impact on training dynamics and optimization. Due to the limitations posed by standard training practices, we propose a tailored initialization scheme and identify important learning rate requirements necessary for training factorized networks. We demonstrate the effectiveness of our deep weight factorization through experiments on various architectures and datasets, consistently outperforming its shallow counterpart and widely used pruning methods.

Fine-tuned large language models (LLMs) often exhibit overconfidence, particularly when trained on small datasets, resulting in poor calibration and inaccurate uncertainty estimates. Evidential Deep Learning (EDL), an uncertainty-aware approach, enables uncertainty estimation in a single forward pass, making it a promising method for calibrating fine-tuned LLMs. However, despite its computational efficiency, EDL is prone to overfitting, as its training objective can result in overly concentrated probability distributions. To mitigate this, we propose regularizing EDL by incorporating an information bottleneck (IB). Our approach IB-EDL suppresses spurious information in the evidence generated by the model and encourages truly predictive information to influence both the predictions and uncertainty estimates. Extensive experiments across various fine-tuned LLMs and tasks demonstrate that IB-EDL outperforms both existing EDL and non-EDL approaches. By improving the trustworthiness of LLMs, IB-EDL facilitates their broader adoption in domains requiring high levels of confidence calibration.

Recent years have witnessed growing interest in semi-implicit variational inference (SIVI) methods due to their ability to rapidly generate samples from complex distributions. However, since the likelihood of these samples is non-trivial to estimate in high dimensions, current research focuses on finding effective SIVI training routines. Although unbiased implicit variational inference (UIVI) has largely been dismissed as imprecise and computationally prohibitive because of its inner MCMC loop, we revisit this method and show that UIVI's MCMC loop can be effectively replaced via importance sampling and the optimal proposal distribution can be learned stably by minimizing an expected forward Kullback–Leibler divergence without bias. Our refined approach demonstrates superior performance or parity with state-of-the-art methods on established SIVI benchmarks.

Structured sparsity regularization offers a principled way to compact neural networks, but its non-differentiability breaks compatibility with conventional stochastic gradient descent and requires either specialized optimizers or additional post-hoc pruning without formal guarantees. In this work, we propose $D$-Gating, a fully differentiable structured overparameterization that splits each group of weights into a primary weight vector and multiple scalar gating factors. We prove that any local minimum under $D$-Gating is also a local minimum using non-smooth structured $L_{2,2/D}$ penalization, and further show that the $D$-Gating objective converges at least exponentially fast to the $L_{2,2/D}$–regularized loss in the gradient flow limit. Together, our results show that $D$-Gating is theoretically equivalent to solving the original group sparsity problem, yet induces distinct learning dynamics that evolve from a non-sparse regime into sparse optimization. We validate our theory across vision, language, and tabular tasks, where $D$-Gating consistently delivers strong performance–sparsity tradeoffs and outperforms both direct optimization of structured penalties and conventional pruning baselines.

% Self-supervised learning methods have shown promising results across a wide range of tasks in computer vision, natural language processing, and multimodal analysis. However, self-supervised approaches come with a notable limitation, dimensional collapse, where a model doesn't fully utilize its capacity to encode information optimally. Motivated by this, we propose ProSMin, a novel probabilistic self-supervised learning approach that leverages the power of probabilistic models to enhance representation quality and mitigate collapsing representations. Our proposed approach involves two neural networks, the online network and the target network, which collaborate and learn the diverse distribution of representations from each other through probabilistic knowledge distillation. The two networks are trained via our new loss function based on proper scoring rules. We provide a theoretical justification for ProSMin and demonstrate its modified scoring rule. This insight validates the method's optimization process and contributes to its robustness and effectiveness in improving representation quality. We evaluate our probabilistic model on various downstream tasks, such as in-distribution generalization, out-of-distribution detection, dataset corruption, low-shot learning, and transfer learning. Our method achieves superior accuracy and calibration, outperforming the self-supervised baseline in a variety of experiments on large datasets such as ImageNet-O and ImageNet-C. ProSMin thus demonstrates its scalability and real-world applicability. Our code is publicly available: https://github.com/amirvhd/SSL-sore-rule.

We warn against a common but incomplete understanding of empirical research in machine learning that leads to non-replicable results, makes findings unreliable, and threatens to undermine progress in the field. To overcome this alarming situation, we call for more awareness of the plurality of ways of gaining knowledge experimentally but also of some epistemic limitations. In particular, we argue most current empirical machine learning research is fashioned as confirmatory research while it should rather be considered exploratory.

Automated machine learning (AutoML) was formed around the fundamental objectives of automatically and efficiently configuring machine learning (ML) workflows, aiding the research of new ML algorithms, and contributing to the democratization of ML by making it accessible to a broader audience. Over the past decade, commendable achievements in AutoML have primarily focused on optimizing predictive performance. This focused progress, while substantial, raises questions about how well AutoML has met its broader, original goals. In this position paper, we argue that a key to unlocking AutoML’s full potential lies in addressing the currently underexplored aspect of user interaction with AutoML systems, including their diverse roles, expectations, and expertise. We envision a more human-centered approach in future AutoML research, promoting the collaborative design of ML systems that tightly integrates the complementary strengths of human expertise and AutoML methodologies.

A major challenge in sample-based inference (SBI) for Bayesian neural networks is the size and structure of the networks’ parameter space. Our work shows that successful SBI is possible by embracing the characteristic relationship between weight and function space, uncovering a systematic link between overparameterization and the difficulty of the sampling problem. Through extensive experiments, we establish practical guidelines for sampling and convergence diagnosis. As a result, we present a deep ensemble initialized approach as an effective solution with competitive performance and uncertainty quantification.

Hyperparameter optimization is crucial for obtaining peak performance of machine learning models. The standard protocol evaluates various hyperparameter configurations using a resampling estimate of the generalization error to guide optimization and select a final hyperparameter configuration. Without much evidence, paired resampling splits, i.e., either a fixed train-validation split or a fixed cross-validation scheme, are often recommended. We show that, surprisingly, reshuffling the splits for every configuration often improves the final model's generalization performance on unseen data. Our theoretical analysis explains how reshuffling affects the asymptotic behavior of the validation loss surface and provides a bound on the expected regret in the limiting regime. This bound connects the potential benefits of reshuffling to the signal and noise characteristics of the underlying optimization problem. We confirm our theoretical results in a controlled simulation study and demonstrate the practical usefulness of reshuffling in a large-scale, realistic hyperparameter optimization experiment. While reshuffling leads to test performances that are competitive with using fixed splits, it drastically improves results for a single train-validation holdout protocol and can often make holdout become competitive with standard CV while being computationally cheaper.

Undersampling is a common method in Magnetic Resonance Imaging (MRI) to subsample the number of data points in k-space, reducing acquisition times at the cost of decreased image quality. A popular approach is to employ undersampling patterns following various strategies, e.g., variable density sampling or radial trajectories. In this work, we propose a method that directly learns the under-sampling masks from data points, thereby also providing task- and domain-specific patterns. To solve the resulting discrete optimization problem, we propose a general optimization routine called ProM: A fully probabilistic, differentiable, versatile, and model-free framework for mask optimization that enforces acceleration factors through a convex constraint. Analyzing knee, brain, and cardiac MRI datasets with our method, we discover that different anatomic regions reveal distinct optimal undersampling masks, demonstrating the benefits of using custom masks, tailored for a downstream task. For example, ProM can create undersampling masks that maximize performance in downstream tasks like segmentation with networks trained on fully-sampled MRIs. Even with extreme acceleration factors, ProM yields reasonable performance while being more versatile than existing methods, paving the way for data-driven all-purpose mask generation.1

Abstract The interpretation of feature importance in machine learning models is challenging when features are dependent. Permutation feature importance (PFI) ignores such dependencies, which can cause misleading interpretations due to extrapolation. A possible remedy is more advanced conditional PFI approaches that enable the assessment of feature importance conditional on all other features. Due to this shift in perspective and in order to enable correct interpretations, it is beneficial if the conditioning is transparent and comprehensible. In this paper, we propose a new sampling mechanism for the conditional distribution based on permutations in conditional subgroups. As these subgroups are constructed using tree-based methods such as transformation trees, the conditioning becomes inherently interpretable. This not only provides a simple and effective estimator of conditional PFI, but also local PFI estimates within the subgroups. In addition, we apply the conditional subgroups approach to partial dependence plots, a popular method for describing feature effects that can also suffer from extrapolation when features are dependent and interactions are present in the model. In simulations and a real-world application, we demonstrate the advantages of the conditional subgroup approach over existing methods: It allows to compute conditional PFI that is more true to the data than existing proposals and enables a fine-grained interpretation of feature effects and importance within the conditional subgroups.

Abstract Beta coefficients for linear regression models represent the ideal form of an interpretable feature effect. However, for non-linear models such as generalized linear models, the estimated coefficients cannot be interpreted as a direct feature effect on the predicted outcome. Hence, marginal effects are typically used as approximations for feature effects, either as derivatives of the prediction function or forward differences in prediction due to changes in feature values. While marginal effects are commonly used in many scientific fields, they have not yet been adopted as a general model-agnostic interpretation method for machine learning models. This may stem from the ambiguity surrounding marginal effects and their inability to deal with the non-linearities found in black box models. We introduce a unified definition of forward marginal effects (FMEs) that includes univariate and multivariate, as well as continuous, categorical, and mixed-type features. To account for the non-linearity of prediction functions, we introduce a non-linearity measure for FMEs. Furthermore, we argue against summarizing feature effects of a non-linear prediction function in a single metric such as the average marginal effect. Instead, we propose to average homogeneous FMEs within population subgroups, which serve as conditional feature effect estimates.

The field of automated machine learning (AutoML) introduces techniques that automate parts of the development of machine learning (ML) systems, accelerating the process and reducing barriers for novices. However, decisions derived from ML models can reproduce, amplify, or even introduce unfairness in our societies, causing harm to (groups of) individuals. In response, researchers have started to propose AutoML systems that jointly optimize fairness and predictive performance to mitigate fairness-related harm. However, fairness is a complex and inherently interdisciplinary subject, and solely posing it as an optimization problem can have adverse side effects. With this work, we aim to raise awareness among developers of AutoML systems about such limitations of fairness-aware AutoML, while also calling attention to the potential of AutoML as a tool for fairness research. We present a comprehensive overview of different ways in which fairness-related harm can arise and the ensuing implications for the design of fairness-aware AutoML. We conclude that while fairness cannot be automated, fairness-aware AutoML can play an important role in the toolbox of ML practitioners. We highlight several open technical challenges for future work in this direction. Additionally, we advocate for the creation of more user-centered assistive systems designed to tackle challenges encountered in fairness work. This article appears in the AI & Society track.

Comparing different AutoML frameworks is notoriously challenging and often done incorrectly. We introduce an open and extensible benchmark that follows best practices and avoids common mistakes when comparing AutoML frameworks. We conduct a thorough comparison of 9 well-known AutoML frameworks across 71 classification and 33 regression tasks. The differences between the AutoML frameworks are explored with a multi-faceted analysis, evaluating model accuracy, its trade-offs with inference time, and framework failures. We also use Bradley-Terry trees to discover subsets of tasks where the relative AutoML framework rankings differ. The benchmark comes with an open-source tool that integrates with many AutoML frameworks and automates the empirical evaluation process end-to-end: from framework installation and resource allocation to in-depth evaluation. The benchmark uses public data sets, can be easily extended with other AutoML frameworks and tasks, and has a website with up-to-date results.

Global feature effect methods, such as partial dependence plots, provide an intelligible visualization of the expected marginal feature effect. However, such global feature effect methods can be misleading, as they do not represent local feature effects of single observations well when feature interactions are present. We formally introduce generalized additive decomposition of global effects (GADGET), which is a new framework based on recursive partitioning to find interpretable regions in the feature space such that the interaction-related heterogeneity of local feature effects is minimized. We provide a mathematical foundation of the framework and show that it is applicable to the most popular methods to visualize marginal feature effects, namely partial dependence, accumulated local effects, and Shapley additive explanations (SHAP) dependence. Furthermore, we introduce and validate a new permutation-based interaction detection procedure that is applicable to any feature effect method that fits into our proposed framework. We empirically evaluate the theoretical characteristics of the proposed methods based on various feature effect methods in different experimental settings. Moreover, we apply our introduced methodology to three real-world examples to showcase their usefulness.

The localization of objects is essential in many applications, such as robotics, virtual and augmented reality, and warehouse logistics. Recent advancements in deep learning have enabled localization using monocular cameras. Traditionally, structure from motion (SfM) techniques predict an object’s absolute position from a point cloud, while absolute pose regression (APR) methods use neural networks to understand the environment semantically. However, both approaches face challenges from environmental factors like motion blur, lighting changes, repetitive patterns, and featureless areas. This study addresses these challenges by incorporating additional information and refining absolute pose estimates with relative pose regression (RPR) methods. RPR also struggles with issues like motion blur. To overcome this, we compute the optical flow between consecutive images using the Lucas–Kanade algorithm and use a small recurrent convolutional network to predict relative poses. Combining absolute and relative poses is difficult due to differences between global and local coordinate systems. Current methods use pose graph optimization (PGO) to align these poses. In this work, we propose recurrent fusion networks to better integrate absolute and relative pose predictions, enhancing the accuracy of absolute pose estimates. We evaluate eight different recurrent units and create a simulation environment to pre-train the APR and RPR networks for improved generalization. Additionally, we record a large dataset of various scenarios in a challenging indoor environment resembling a warehouse with transportation robots. Through hyperparameter searches and experiments, we demonstrate that our recurrent fusion method outperforms PGO in effectiveness.

Semi-structured regression (SSR) models jointly learn the effect of structured (tabular) and unstructured (non-tabular) data through additive predictors and deep neural networks (DNNs), respectively. Inference in SSR models aims at deriving confidence intervals for the structured predictor, although current approaches ignore the variance of the DNN estimation of the unstructured effects. This results in an underestimation of the variance of the structured coefficients and, thus, an increase of Type-I error rates. To address this shortcoming, we present here a theoretical framework for structured inference in SSR models that incorporates the variance of the DNN estimate into confidence intervals for the structured predictor. By treating this estimate as a random offset with known variance, our formulation is agnostic to the specific deep uncertainty quantification method employed. Through numerical experiments and a practical application on a medical dataset, we show that our approach results in increased coverage of the true structured coefficients and thus a reduction in Type-I error rate compared to ignoring the variance of the neural network, naive ensembling of SSR models, and a variational inference baseline.

Deep learning has enabled outstanding progress on bioinformatics datasets and a variety of tasks, such as protein structure prediction, identification of regulatory regions, genome annotation, and interpretation of the noncoding genome. The layout and configuration of neural networks used for these tasks have mostly been developed manually by human experts, which is a time-consuming and error-prone process. Therefore, there is growing interest in automated neural architecture search (NAS) methods in bioinformatics. In this paper, we present a novel search space for NAS algorithms that operate on genome data, thus creating extensions for existing NAS algorithms for sequence data that we name Genome-DARTS, Genome-P-DARTS, Genome-BONAS, Genome-SH, and Genome-RS. Moreover, we introduce two novel NAS algorithms, CWP-DARTS and EDPDARTS, that build on and extend the idea of P-DARTS. We evaluate the presented methods and compare them to manually designed neural architectures on a widely used genome sequence machine learning task to show that NAS methods can be adapted well for bioinformatics sequence datasets. Our experiments show that architectures optimized by our NAS methods outperform manually developed architectures while having significantly fewer parameters.

Artificial benchmark functions are commonly used in optimization research because of their ability to rapidly evaluate potential solutions, making them a preferred substitute for real-world problems. However, these benchmark functions have faced criticism for their limited resemblance to real-world problems. In response, recent research has focused on automatically generating new benchmark functions for areas where established test suites are inadequate. These approaches have limitations, such as the difficulty of generating new benchmark functions that exhibit exploratory landscape analysis (ELA) features beyond those of existing benchmarks.

We present a model-agnostic framework for jointly optimizing the predictive performance and interpretability of supervised machine learning models for tabular data. Interpretability is quantified via three measures: feature sparsity, interaction sparsity of features, and sparsity of non-monotone feature effects. By treating hyperparameter optimization of a machine learning algorithm as a multi-objective optimization problem, our framework allows for generating diverse models that trade off high performance and ease of interpretability in a single optimization run. Efficient optimization is achieved via augmentation of the search space of the learning algorithm by incorporating feature selection, interaction and monotonicity constraints into the hyperparameter search space. We demonstrate that the optimization problem effectively translates to finding the Pareto optimal set of groups of selected features that are allowed to interact in a model, along with finding their optimal monotonicity constraints and optimal hyperparameters of the learning algorithm itself. We then introduce a novel evolutionary algorithm that can operate efficiently on this augmented search space. In benchmark experiments, we show that our framework is capable of finding diverse models that are highly competitive or outperform state-of-the-art XGBoost or Explainable Boosting Machine models, both with respect to performance and interpretability.

The field of computational biology has been enhanced by deep learning models, which hold great promise for revolutionizing domains such as protein folding and drug discovery. Recent studies have underscored the tremendous potential of these models, particularly in the realm of gene regulation and the more profound understanding of the non-coding regions of the genome. On the other hand, this raises significant concerns about the reliability and efficacy of such models, which have their own biases by design, along with those learned from the data. Uncertainty quantification allows us to measure where the system is confident and know when it can be trusted. In this paper, we study several uncertainty quantification methods with respect to a multi-target regression task, specifically predicting regulatory activity profiles using DNA sequence data. Using the Basenji model, we investigate how such methods can improve in-domain generalization, out-of-distribution detection, and provide coverage guarantees on prediction intervals.

Constrained clustering allows the training of classi-fication models using pairwise constraints only, which are weak and relatively easy to mine, while still yielding full-supervision-level model performance. While they perform well even in the absence of the true underlying class labels, constrained clustering models still require large amounts of binary constraint annotations for training. In this paper, we propose a semi-supervised context whereby a large amount of unconstrained data is available alongside a smaller set of constraints, and propose ConstraintMatch to leverage such unconstrained data. While a great deal of progress has been made in semi-supervised learning using full labels, there are a number of challenges that prevent a naive application of the resulting methods in the constraint-based label setting. Therefore, we reason about and analyze these challenges, specifically 1) proposing a pseudo-constraining mechanism to overcome the confirmation bias, a major weakness of pseudo-labeling, 2) developing new methods for pseudo-labeling towards the selection of informative unconstrained samples, 3) showing that this also allows the use of pairwise loss functions for the initial and auxiliary losses which facilitates semi-constrained model training. In extensive experiments, we demonstrate the effectiveness of ConstraintMatch over relevant baselines in both the regular clustering and overclustering scenarios on five challenging benchmarks and provide analyses of its several components.

The quantification of aleatoric and epistemic uncertainty in terms of conditional entropy and mutual information, respectively, has recently become quite common in machine learning. While the properties of these measures, which are rooted in information theory, seem appealing at first glance, we identify various incoherencies that call their appropriateness into question. In addition to the measures themselves, we critically discuss the idea of an additive decomposition of total uncertainty into its aleatoric and epistemic constituents. Experiments across different computer vision tasks support our theoretical findings and raise concerns about current practice in uncertainty quantification.

Hyperparameter optimization constitutes a large part of typical modern machine learning (ML) workflows. This arises from the fact that ML methods and corresponding preprocessing steps often only yield optimal performance when hyperparameters are properly tuned. But in many applications, we are not only interested in optimizing ML pipelines solely for predictive accuracy; additional metrics or constraints must be considered when determining an optimal configuration, resulting in a multi-objective optimization problem. This is often neglected in practice, due to a lack of knowledge and readily available software implementations for multi-objective hyperparameter optimization. In this work, we introduce the reader to the basics of multi-objective hyperparameter optimization and motivate its usefulness in applied ML. Furthermore, we provide an extensive survey of existing optimization strategies from the domains of evolutionary algorithms and Bayesian optimization. We illustrate the utility of multi-objective optimization in several specific ML applications, considering objectives such as operating conditions, prediction time, sparseness, fairness, interpretability, and robustness.

Machine learning models can automatically learn complex relationships, such as non-linear and interaction effects. Interpretable machine learning methods such as partial dependence plots visualize marginal feature effects but may lead to misleading interpretations when feature interactions are present. Hence, employing additional methods that can detect and measure the strength of interactions is paramount to better understand the inner workings of machine learning models. We demonstrate several drawbacks of existing global interaction detection approaches, characterize them theoretically, and evaluate them empirically. Furthermore, we introduce regional effect plots with implicit interaction detection, a novel framework to detect interactions between a feature of interest and other features. The framework also quantifies the strength of interactions and provides interpretable and distinct regions in which feature effects can be interpreted more reliably, as they are less confounded by interactions. We prove the theoretical eligibility of our method and show its applicability on various simulation and real-world examples.

The performance of a machine learning model degrades when it is applied to data from a similar but different domain than the data it has initially been trained on. To mitigate this domain shift problem, domain adaptation (DA) techniques search for an optimal transformation that converts the (current) input data from a source domain to a target domain to learn a domain-invariant representation that reduces domain discrepancy. This paper proposes a novel supervised DA based on two steps. First, we search for an optimal class-dependent transformation from the source to the target domain from a few samples. We consider optimal transport methods such as the earth mover's distance, Sinkhorn transport and correlation alignment. Second, we use embedding similarity techniques to select the corresponding transformation at inference. We use correlation metrics and higher-order moment matching techniques. We conduct an extensive evaluation on time-series datasets with domain shift including simulated and various online handwriting datasets to demonstrate the performance.

The ability to estimate epistemic uncertainty is often crucial when deploying machine learning in the real world, but modern methods often produce overconfident, uncalibrated uncertainty predictions. A common approach to quantify epistemic uncertainty, usable across a wide class of prediction models, is to train a model ensemble. In a naive implementation, the ensemble approach has high computational cost and high memory demand. This challenges in particular modern deep learning, where even a single deep network is already demanding in terms of compute and memory, and has given rise to a number of attempts to emulate the model ensemble without actually instantiating separate ensemble members. We introduce FiLM-Ensemble, a deep, implicit ensemble method based on the concept of Feature-wise Linear Modulation (FiLM). That technique was originally developed for multi-task learning, with the aim of decoupling different tasks. We show that the idea can be extended to uncertainty quantification: by modulating the network activations of a single deep network with FiLM, one obtains a model ensemble with high diversity, and consequently well-calibrated estimates of epistemic uncertainty, with low computational overhead in comparison. Empirically, FiLM-Ensemble outperforms other implicit ensemble methods, and it comes very close to the upper bound of an explicit ensemble of networks (sometimes even beating it), at a fraction of the memory cost.

Abstract Hyperparameter optimization (HPO) is a key component of machine learning models for achieving peak predictive performance. While numerous methods and algorithms for HPO have been proposed over the last years, little progress has been made in illuminating and examining the actual structure of these black-box optimization problems. Exploratory landscape analysis (ELA) subsumes a set of techniques that can be used to gain knowledge about properties of unknown optimization problems. In this paper, we evaluate the performance of five different black-box optimizers on 30 HPO problems, which consist of two-, three- and five-dimensional continuous search spaces of the XGBoost learner trained on 10 different data sets. This is contrasted with the performance of the same optimizers evaluated on 360 problem instances from the black-box optimization benchmark (BBOB). We then compute ELA features on the HPO and BBOB problems and examine similarities and differences. A cluster analysis of the HPO and BBOB problems in ELA feature space allows us to identify how the HPO problems compare to the BBOB problems on a structural meta-level. We identify a subset of BBOB problems that are close to the HPO problems in ELA feature space and show that optimizer performance is comparably similar on these two sets of benchmark problems. We highlight open challenges of ELA for HPO and discuss potential directions of future research and applications.

Multivariate Time Series (MTS) classification is important in various applications such as signature verification, person identification, and motion recognition. In deep learning these classification tasks are usually learned using the cross-entropy loss. A related yet different task is predicting trajectories observed as MTS. Important use cases include handwriting reconstruction, shape analysis, and human pose estimation. The goal is to align an arbitrary dimensional time series with its ground truth as accurately as possible while reducing the error in the prediction with a distance loss and the variance with a similarity loss. Although learning both losses with Multi-Task Learning (MTL) helps to improve trajectory alignment, learning often remains difficult as both tasks are contradictory. We propose a novel neural network architecture for MTL that notably improves the MTS classification and trajectory regression performance in online handwriting (OnHW) recognition. We achieve this by jointly learning the cross-entropy loss in combination with distance and similarity losses. On an OnHW task of handwritten characters with multivariate inertial and visual data inputs we are able to achieve crucial improvements (lower error with less variance) of trajectory prediction while still improving the character classification accuracy in comparison to models trained on the individual tasks.

Abstract Interpretable machine learning has become a very active area of research due to the rising popularity of machine learning algorithms and their inherently challenging interpretability. Most work in this area has been focused on the interpretation of single features in a model. However, for researchers and practitioners, it is often equally important to quantify the importance or visualize the effect of feature groups. To address this research gap, we provide a comprehensive overview of how existing model-agnostic techniques can be defined for feature groups to assess the grouped feature importance, focusing on permutation-based, refitting, and Shapley-based methods. We also introduce an importance-based sequential procedure that identifies a stable and well-performing combination of features in the grouped feature space. Furthermore, we introduce the combined features effect plot, which is a technique to visualize the effect of a group of features based on a sparse, interpretable linear combination of features. We used simulation studies and real data examples to analyze, compare, and discuss these methods.

Abstract Handwriting is one of the most frequently occurring patterns in everyday life and with it comes challenging applications such as handwriting recognition, writer identification and signature verification. In contrast to offline HWR that only uses spatial information (i.e., images), online HWR uses richer spatio-temporal information (i.e., trajectory data or inertial data). While there exist many offline HWR datasets, there are only little data available for the development of OnHWR methods on paper as it requires hardware-integrated pens. This paper presents data and benchmark models for real-time sequence-to-sequence learning and single character-based recognition. Our data are recorded by a sensor-enhanced ballpoint pen, yielding sensor data streams from triaxial accelerometers, a gyroscope, a magnetometer and a force sensor at 100 Hz. We propose a variety of datasets including equations and words for both the writer-dependent and writer-independent tasks. Our datasets allow a comparison between classical OnHWR on tablets and on paper with sensor-enhanced pens. We provide an evaluation benchmark for seq2seq and single character-based HWR using recurrent and temporal convolutional networks and transformers combined with a connectionist temporal classification (CTC) loss and cross-entropy (CE) losses. Our convolutional network combined with BiLSTMs outperforms transformer-based architectures, is on par with InceptionTime for sequence-based classification tasks and yields better results compared to 28 state-of-the-art techniques. Time-series augmentation methods improve the sequence-based task, and we show that CE variants can improve the single classification task. Our implementations together with the large benchmark of state-of-the-art techniques of novel OnHWR datasets serve as a baseline for future research in the area of OnHWR on paper.

Automated hyperparameter optimization (HPO) has gained great popularity and is an important component of most automated machine learning frameworks. However, the process of designing HPO algorithms is still an unsystematic and manual process: new algorithms are often built on top of prior work, where limitations are identified and improvements are proposed. Even though this approach is guided by expert knowledge, it is still somewhat arbitrary. The process rarely allows for gaining a holistic understanding of which algorithmic components drive performance and carries the risk of overlooking good algorithmic design choices. We present a principled approach to automated benchmark-driven algorithm design applied to multifidelity HPO (MF-HPO). First, we formalize a rich space of MF-HPO candidates that includes, but is not limited to, common existing HPO algorithms and then present a configurable framework covering this space. To find the best candidate automatically and systematically, we follow a programming-by-optimization approach and search over the space of algorithm candidates via Bayesian optimization. We challenge whether the found design choices are necessary or could be replaced by more naive and simpler ones by performing an ablation analysis. We observe that using a relatively simple configuration (in some ways, simpler than established methods) performs very well as long as some critical configuration parameters are set to the right value.

While deep semi-supervised learning has gained much attention in computer vision, limited research exists on its applicability in the time series domain. In this work, we investigate the transferability of state-of-the-art deep semi-supervised models from image to time series classification. We discuss the necessary model adaptations, in particular an appropriate model backbone architecture and the use of tailored data augmentation strategies. Based on these adaptations, we explore the potential of deep semi-supervised learning in the context of time series classification by evaluating our methods on large public time series classification problems with varying amounts of labeled samples. We perform extensive comparisons under a decidedly realistic and appropriate evaluation scheme with a unified reimplementation of all algorithms considered, which is yet lacking in the field. We find that these transferred semi-supervised models show significant performance gains over strong supervised, semi-supervised and self-supervised alternatives, especially for scenarios with very few labeled samples.

Automated hyperparameter optimization (HPO) can support practitioners to obtain peak performance in machine learning models.However, there is often a lack of valuable insights into the effects of different hyperparameters on the final model performance.This lack of explainability makes it difficult to trust and understand the automated HPO process and its results.We suggest using interpretable machine learning (IML) to gain insights from the experimental data obtained during HPO with Bayesian optimization (BO).BO tends to focus on promising regions with potential high-performance configurations and thus induces a sampling bias.Hence, many IML techniques, such as the partial dependence plot (PDP), carry the risk of generating biased interpretations.By leveraging the posterior uncertainty of the BO surrogate model, we introduce a variant of the PDP with estimated confidence bands.We propose to partition the hyperparameter space to obtain more confident and reliable PDPs in relevant sub-regions.In an experimental study, we provide quantitative evidence for the increased quality of the PDPs within sub-regions.

SUMMARY: As machine learning has become increasingly popular over the last few decades, so too has the number of machine-learning interfaces for implementing these models. Whilst many R libraries exist for machine learning, very few offer extended support for survival analysis. This is problematic considering its importance in fields like medicine, bioinformatics, economics, engineering and more. mlr3proba provides a comprehensive machine-learning interface for survival analysis and connects with mlr3's general model tuning and benchmarking facilities to provide a systematic infrastructure for survival modelling and evaluation. AVAILABILITY AND IMPLEMENTATION: mlr3proba is available under an LGPL-3 licence on CRAN and at https://github.com/mlr-org/mlr3proba, with further documentation at https://mlr3book.mlr-org.com/survival.html.

Recent years have seen a proliferation of ML frameworks. Such systems make ML accessible to non-experts, especially when combined with powerful parameter tuning and AutoML techniques. Modern, applied ML extends beyond direct learning on clean data, however, and needs an expressive language for the construction of complex ML workflows beyond simple pre- and post-processing. We present mlr3pipelines, an R framework which can be used to define linear and complex non-linear ML workflows as directed acyclic graphs. The framework is part of the mlr3 ecosystem, leveraging convenient resampling, benchmarking, and tuning components.

Both feature selection and hyperparameter tuning are key tasks in machine learning. Hyperparameter tuning is often useful to increase model performance, while feature selection is undertaken to attain sparse models. Sparsity may yield better model interpretability and lower cost of data acquisition, data handling and model inference. While sparsity may have a beneficial or detrimental effect on predictive performance, a small drop in performance may be acceptable in return for a substantial gain in sparseness. We therefore treat feature selection as a multi-objective optimization task. We perform hyperparameter tuning and feature selection simultaneously because the choice of features of a model may influence what hyperparameters perform well.

Interpretable Machine Learning (IML) methods are used to gain insight into the relevance of a feature of interest for the performance of a model. Commonly used IML methods differ in whether they consider features of interest in isolation, e.g., Permutation Feature Importance (PFI), or in relation to all remaining feature variables, e.g., Conditional Feature Importance (CFI). As such, the perturbation mechanisms inherent to PFI and CFI represent extreme reference points. We introduce Relative Feature Importance (RFI), a generalization of PFI and CFI that allows for a more nuanced feature importance computation beyond the PFI versus CFI dichotomy. With RFI, the importance of a feature relative to any other subset of features can be assessed, including variables that were not available at training time. We derive general interpretation rules for RFI based on a detailed theoretical analysis of the implications of relative feature relevance, and demonstrate the method's usefulness on simulated examples.

Counterfactual explanations are one of the most popular methods to make predictions of black box machine learning models interpretable by providing explanations in the form of ‘what-if scenarios’. Most current approaches optimize a collapsed, weighted sum of multiple objectives, which are naturally difficult to balance a-priori. We propose the Multi-Objective Counterfactuals (MOC) method, which translates the counterfactual search into a multi-objective optimization problem. Our approach not only returns a diverse set of counterfactuals with different trade-offs between the proposed objectives, but also maintains diversity in feature space. This enables a more detailed post-hoc analysis to facilitate better understanding and also more options for actionable user responses to change the predicted outcome. Our approach is also model-agnostic and works for numerical and categorical input features. We show the usefulness of MOC in concrete cases and compare our approach with state-of-the-art methods for counterfactual explanations.

Modern supervised machine learning algorithms involve hyperparameters that have to be set before running them. Options for setting hyperparameters are default values from the software package, manual configuration by the user or configuring them for optimal predictive performance by a tuning procedure. The goal of this paper is two-fold. Firstly, we formalize the problem of tuning from a statistical point of view, define data-based defaults and suggest general measures quantifying the tunability of hyperparameters of algorithms. Secondly, we conduct a large-scale benchmarking study based on 38 datasets from the OpenML platform and six common machine learning algorithms. We apply our measures to assess the tunability of their parameters. Our results yield default values for hyperparameters and enable users to decide whether it is worth conducting a possibly time consuming tuning strategy, to focus on the most important hyperparameters and to choose adequate hyperparameter spaces for tuning.

Algorithm selection approaches have achieved impressive performance improvements in many areas of AI. Most of the literature considers the offline algorithm selection problem, where the initial selection model is never updated after training. However, new data from running algorithms on instances becomes available when algorithms are selected and run. We investigate how this online data can be used to improve the selection model over time. This is especially relevant when insufficient training instances were used, but potentially improves the performance of algorithm selection in all cases. We formally define the online algorithm selection problem and model it as a contextual multi-armed bandit problem, propose a methodology for solving it, and empirically demonstrate performance improvements. We also show that our online algorithm selection method can be used when no training data whatsoever is available, a setting where offline algorithm selection cannot be used. Our experiments indicate that a simple greedy approach achieves the best performance.

The mlr package provides a generic, object- oriented, and extensible framework for classification, regression, survival analysis and clustering for the R language. It provides a unified interface to more than 160 basic learners and includes meta-algorithms and model selection techniques to improve and extend the functionality of basic learners with, e.g., hyperparameter tuning, feature selection, and ensemble construction. Parallel high-performance computing is natively supported. The package targets practitioners who want to quickly apply machine learning algorithms, as well as researchers who want to implement, benchmark, and compare their new methods in a structured environment.

The majority of algorithms can be controlled or adjusted by parameters. Their values can substantially affect the algorithms' performance. Since the manual exploration of the parameter space is tedious -- even for few parameters -- several automatic procedures for parameter tuning have been proposed. Recent approaches also take into account some characteristic properties of the problem instances, frequently termed instance features.

Many classification algorithms, such as Neural Networks and Support Vector Machines, have a range of hyper-parameters that may strongly affect the predictive performance of the models induced by them. Hence, it is recommended to define the values of these hyper-parameters using optimization techniques. While these techniques usually converge to a good set of values, they typically have a high computational cost, because many candidate sets of values are evaluated during the optimization process. It is often not clear whether this will result in parameter settings that are significantly better than the default settings. When training time is limited, it may help to know when these parameters should definitely be tuned. In this study, we use meta-learning to predict when optimization techniques are expected to lead to models whose predictive performance is better than those obtained by using default parameter settings. Hence, we can choose to employ optimization techniques only when they are expected to improve performance, thus reducing the overall computational cost. We evaluate these meta-learning techniques on more than one hundred data sets. The experimental results show that it is possible to accurately predict when optimization techniques should be used instead of default values suggested by some machine learning libraries.

Classification is one of the most common machine learning tasks. SVMs have been frequently applied to this task. In general, the values chosen for the hyper-parameters of SVMs affect the performance of their induced predictive models. Several studies use optimization techniques to find a set of hyper-parameter values that induces classifiers with good predictive performance. This paper investigates the hypothesis that a simple Random Search method is sufficient to adjust the hyper-parameters of SVMs. A set of experiments compared the performance of five tuning techniques: three meta-heuristics commonly used, Random Search and Grid Search. The experimental results show that the predictive performance of models using Random Search is equivalent to those obtained using meta-heuristics and Grid Search, but with a lower computational cost.

We present methods to answer two basic questions that arise when benchmarking optimization algorithms. The first one is: which algorithm is the "best" one? and the second one is: which algorithm should I use for my real-world problem? Both are connected and neither is easy to answer. We present a theoretical framework for designing and analyzing the raw data of such benchmark experiments. This represents a first step in answering the aforementioned questions. The 2009 and 2010 BBOB benchmark results are analyzed by means of this framework and we derive insight regarding the answers to the two questions. Furthermore, we discuss how to properly aggregate rankings from algorithm evaluations on individual problems into a consensus, its theoretical background and which common pitfalls should be avoided. Finally, we address the grouping of test problems into sets with similar optimizer rankings and investigate whether these are reflected by already proposed test problem characteristics, finding that this is not always the case.

Understanding the behaviour of well-known algorithms for classical NP-hard optimisation problems is still a difficult task. With this paper, we contribute to this research direction and carry out a feature based comparison of local search and the well-known Christofides approximation algorithm for the Traveling Salesperson Problem. We use an evolutionary algorithm approach to construct easy and hard instances for the Christofides algorithm, where we measure hardness in terms of approximation ratio. Our results point out important features and lead to hard and easy instances for this famous algorithm. Furthermore, our cross-comparison gives new insights on the complementary benefits of the different approaches.

The steady supply of new optimization methods makes the algorithm selection problem (ASP) an increasingly pressing and challenging task, specially for real-world black-box optimization problems. The introduced approach considers the ASP as a cost-sensitive classification task which is based on Exploratory Landscape Analysis. Low-level features gathered by systematic sampling of the function on the feasible set are used to predict a well-performing algorithm out of a given portfolio. Example-specific label costs are defined by the expected runtime of each candidate algorithm. We use one-sided support vector regression to solve this learning problem. The approach is illustrated by means of the optimization problems and algorithms of the BBOB'09/10 workshop.

Meta-modeling has become a crucial tool in solving expensive optimization problems. Much of the work in the past has focused on finding a good regression method to model the fitness function. Examples include classical linear regression, splines, neural networks, Kriging and support vector regression. This paper specifically draws attention to the fact that assessing model accuracy is a crucial aspect in the meta-modeling framework. Resampling strategies such as cross-validation, subsampling, bootstrapping, and nested resampling are prominent methods for model validation and are systematically discussed with respect to possible pitfalls, shortcomings, and specific features. A survey of meta-modeling techniques within evolutionary optimization is provided. In addition, practical examples illustrating some of the pitfalls associated with model selection and performance assessment are presented. Finally, recommendations are given for choosing a model validation technique for a particular setting.

Exploratory Landscape Analysis subsumes a number of techniques employed to obtain knowledge about the properties of an unknown optimization problem, especially insofar as these properties are important for the performance of optimization algorithms. Where in a first attempt, one could rely on high-level features designed by experts, we approach the problem from a different angle here, namely by using relatively cheap low-level computer generated features. Interestingly, very few features are needed to separate the BBOB problem groups and also for relating a problem to high-level, expert designed features, paving the way for automatic algorithm selection.