Model-agnostic interpretation methods can mislead if not used carefully.
problem Misinterpretation of machine learning models due to improper use of techniques.
method General pitfalls of model-agnostic interpretation methods.
result Many pitfalls exist when using global interpretation techniques for machine learning models.
This document provides a tutorial description of the use of the MDL principle in complex graph analysis. We give a brief summary of the preliminary subjects, and describe the basic principle, using the example of analysing the size of the largest clique in a graph. We also provide a discussion of how to interpret the r…
New categorization of community detection methods to avoid pitfalls.
problem Misuse of community detection methods in practice.
method Descriptive vs. inferential approaches.
result Inferential methods provide clearer insights into network formation.
Machine learning suffers from poor design, data, and evaluation practices.
problem Ad hoc design, poor data hygiene, and lack of statistical rigor in model evaluation.
method Examines the entire machine learning process from design to evaluation, highlighting common pitfalls and providing recommendations.
result Common pitfalls in machine learning research and development are identified and actionable recommendations are provided.
This paper evaluates metrics for graph generative models, addressing common pitfalls.
problem Evaluating and comparing graph generative models effectively.
method Systematic evaluation of MMD, analysis of synthetic and real graphs, practical recommendations.
result MMD can be problematic; practical solutions are provided.
Machine learning methods have gained a great deal of popularity in recent years among public administration scholars and practitioners. These techniques open the door to the analysis of text, image and other types of data that allow us to test foundational theories of public administration and to develop new theories. …
Uncertainty estimation and ensembling methods go hand-in-hand. Uncertainty estimation is one of the main benchmarks for assessment of ensembling performance. At the same time, deep learning ensembles have provided state-of-the-art results in uncertainty estimation. In this work, we focus on in-domain uncertainty for im…
Machine Learning (ML) and Deep Learning (DL) innovations are being introduced at such a rapid pace that model owners and evaluators are hard-pressed analyzing and studying them. This is exacerbated by the complicated procedures for evaluation. The lack of standard systems and efficient techniques for specifying and pro…
Proposes a VAE with a discrete bottleneck for better text generation.
problem VAEs struggle with latent variable auto-regressive decoding in text generation.
method Introduces a discretized bottleneck to enforce latent feature matching in a compact space.
result Demonstrates improved text generation capabilities across various tasks.
Study identifies pitfalls in assessing hierarchies for multi-class classification.
problem Lack of understanding in selecting hierarchies for multi-class classification.
method Analyzed and compared popular approaches to extracting hierarchies.
result Hierarchy quality becomes irrelevant when using powerful classifiers.
Benchmark assesses LLMs' causal inference skills, revealing significant limitations.
problem Lack of rigorous evaluation of LLMs' causal inference capabilities.
method CausalPitfalls benchmark with structured challenges and grading rubrics.
result Significant limitations in current LLMs' statistical causal inference.
NESYM combines AI and Earth models for new climate insights.
problem Replacing traditional Earth models with AI.
method Neural Earth System Modelling (NESYM) integrating AI and climate models.
result Artificial intelligence may render traditional models obsolete.
Bayesian optimization improves molecule design by addressing three pitfalls.
problem Bayesian optimization pitfalls cause poor performance in molecule design.
method Identified and addressed three pitfalls: incorrect prior width, over-smoothing, and inadequate acquisition function maximization.
result Basic BO setup achieves highest performance on PMO benchmark.
Develops methods for causal inference in compositional data using instrumental variables.
problem Interpreting summary statistics like diversity indices as causal effects in compositional data.
method Statistical data transformations and regression techniques tailored for compositional data.
result Advantages and limitations of the proposed methods demonstrated on synthetic and real microbiome data.
Unified framework for ranking-and-selection with multiple correct answers and non-answerable estimates
problem Fixed-precision ranking-and-selection in structured settings with non-unique answers and non-answerable estimates
method Unified framework based on answer-wise acceptance sets, restricted generalized likelihood ratio stopping, and answer-pitfall decomposition
result Unified recipe performs well across a broad range of pure-exploration problems
New method corrects correlation bias in feature importance.
problem Correlations between features in high-dimensional data bias statistical and machine learning models.
method Pairwise Permutation Algorithm (PPA)
result PPA corrects correlation bias, identifying biological relevant biomarkers.
Nonnegative matrix factorization (NMF) has become a workhorse for signal and data analytics, triggered by its model parsimony and interpretability. Perhaps a bit surprisingly, the understanding to its model identifiability---the major reason behind the interpretability in many applications such as topic mining and hype…
This paper improves conditional sampling for VAEs by overcoming structural issues.
problem Computational intractability of conditional sampling in VAEs.
method Proposes two methods to address pitfalls in Metropolis-within-Gibbs (MWG) for VAEs.
result Improved performance on sampling tasks.
Recent progress in the field of artificial intelligence, machine learning and also in computer industry resulted in the ongoing boom of using these techniques as applied to solving complex tasks in both science and industry. Same is, of course, true for the financial industry and mathematical finance. In this paper we …
FaIRGP model improves climate emulation with physical interpretability.
problem Lack of physical interpretability in data-driven emulators.
method Bayesian approach to a data-driven emulator of energy balance equations.
result Demonstrates skillful emulation of global and spatial surface temperatures.
A new modeling framework CSN simplifies and interprets machine learning models.
problem Complexity and interpretability issues in machine learning models.
method Combines spline transformation and cross-network to create CSN.
result CSN is as performant and interpretable as XGBoost and FCNN.
This paper evaluates test selection methods for deep neural networks, revealing their limitations.
problem Challenges in testing deep learning systems due to high labeling costs.
method Analysis and empirical testing of 11 test selection methods on five datasets.
result Test selection methods can fail under certain conditions, leading to significant drops in test relative coverage.
Deep learning applied to SAR data is explored in this paper.
problem Limited use of deep learning in SAR data processing.
method Introduction to relevant deep learning models, analysis of SAR data characteristics, review of state-of-the-art applications, and future research directions.
result Unlocking the potential of deep learning in SAR data processing.
This paper highlights the overlooked role of preprocessing hyperparameters in machine learning model performance.
problem The neglect of preprocessing hyperparameters in machine learning model evaluation.
method Empirical review and illustration of different procedures for generating and evaluating prediction models.
result Users may fail to report or account for preprocessing hyperparameter optimization, leading to exaggerated performance claims.
Cross-validation pitfalls in change-point regression are addressed with new approaches.
problem Cross-validation's prediction error-based criterion may lead to under- or over-estimation of change-points.
method Proposes two approaches: absolute error loss and modified holdout sets.
result Consistent estimation of the number of change-points under certain conditions.
AI methods broaden signal discovery in scientific data.
problem Limited coverage of possible signals in model-dependent searches.
method Model-agnostic AI strategies for broad exploration.
result Enhanced discovery potential in experimental science.
This paper improves uncertainty quantification in ELM models.
problem Uncertainty in ELM predictions due to data assumptions and randomness.
method Analytical derivations and variance estimates under various conditions.
result Improved understanding and estimation of ELM variability.
The paper offers a checklist for comparing human and machine visual perception.
problem Comparing human and machine visual perception.
method Designing, conducting, and interpreting experiments to investigate mechanisms.
result Feedback mechanisms may not be necessary for visual reasoning tasks.
New findings show a balance between data fit and complexity in kernel hyperparameters.
problem Overcorrelation due to reparametrization of kernel hyperparameters.
method Reparametrization of kernel hyperparameters and analysis of marginal likelihood.
result Data fit term influences all other kernel hyperparameters, not just the complexity penalty.
LLA shows strong performance in Bayesian optimization but has unbounded search space issues.
problem Applying LLA in unbounded search spaces for Bayesian optimization.
method Linearized-Laplace approximation applied to Bayesian optimization problems.
result LLA demonstrates strong performance but also presents unbounded search space challenges.
Success in the quest for artificial intelligence has the potential to bring unprecedented benefits to humanity, and it is therefore worthwhile to investigate how to maximize these benefits while avoiding potential pitfalls. This article gives numerous examples (which should by no means be construed as an exhaustive lis…
New datasets reveal neural networks can rely on simple features, leading to poor generalization.
problem Neural networks' reliance on simple features can lead to poor generalization and robustness.
method Designing datasets with varying levels of simplicity and incorporating non-robustness.
result Neural networks can exclusively rely on the simplest feature, leading to poor performance on complex data.
Proposes a new undersampling method for imbalanced data classification.
problem Challenges of oversampling and undersampling in imbalanced data.
method Bilevel optimization framework for identifying optimal subset of majority training data.
result Improves F1 scores by up to 10% compared to state-of-the-art methods.
Statistical analysis reveals pitfalls in climate network construction.
problem Uncertainty in correlation estimates affects network characteristics.
method Simulated isotropic random fields and resampling procedures.
result Spurious link bundles and high-degree clusters are common.
TBAL reduces manual annotation but requires validated data.
problem Creating large, high-quality labeled datasets.
method Threshold-based auto-labeling using human validation data.
result Sample complexity bounds on validation data needed.
Statistical tests for fairness in admissions data reveal hidden patterns.
problem Simpson's paradox in admissions data hides true gender bias.
method Introduces a new statistical test based on Pearl's instrumental-variable inequalities.
result Statistical tests for fairness coincide with causal notions for the Berkeley admissions case.
Paper proposes an ensemble of attacks to evaluate adversarial robustness more reliably.
problem Insufficient evaluation of adversarial defenses leads to incorrect robustness assessments.
method Developed two extensions of PGD-attack and combined them with two complementary attacks.
result Identified several broken defenses with lower robust test accuracy than reported.
This document is an invited chapter covering the specificities of ABC model choice, intended for the incoming Handbook of ABC by Sisson, Fan, and Beaumont (2017). Beyond exposing the potential pitfalls of ABC based posterior probabilities, the review emphasizes mostly the solution proposed by Pudlo et al. (2016) on the…
This letter uses the Block Maxima Extreme Value approach to quantify catastrophic risk in international equity markets. Risk measures are generated from a set threshold of the distribution of returns that avoids the pitfall of using absolute returns for markets exhibiting diverging levels of risk. From an application t…
Framework adds human knowledge to AI decisions to improve outcomes.
problem Conflict between AI recommendations and human insights.
method Develops a framework for integrating human knowledge as a guardrail for AI decisions.
result Human knowledge can improve AI decisions, especially in specific pitfalls.
Popular deep learning uncertainty estimation methods often mislead on out-of-distribution data.
problem Misleading uncertainty estimates on out-of-distribution data.
method Analysis of Gaussian process, Bayesian neural networks, and Monte Carlo dropout methods.
result BNNs and MCDropout do not always provide high uncertainty estimates on out-of-distribution samples.
This guide explains statistical distances for evaluating generative models.
problem Evaluating the quality of samples from generative models.
method Four statistical distances: SW, C2ST, MMD, FID.
result Different distances can yield varying results on similar data.
Confirmation bias leads to biased estimates in noisy data analysis.
problem Confirmation bias affects scientific conclusions in noisy data environments.
method Investigation of confirmation bias in Gaussian mixture models using K-means and EM algorithms.
result Estimates from algorithms are biased and resemble initial hypotheses, not the noise.
We construct Zero-Coupon Bond markets driven by a cylindrical Brownian motion in which the notion of generalized portfolio has important flaws: There exist bounded smooth random variables with generalized hedging portfolios for which the price of their risky part is +∞ at each time. For these generalized portfol…
A fundamental problem in computer vision is boundary estimation, where the goal is to delineate the boundary of objects in an image. In this paper, we propose a method which jointly incorporates geometric and topological information within an image to simultaneously estimate boundaries for objects within images with mo…
Comparison of UQ methods in deep learning for a simple physical system.
problem Uncertainty quantification in deep learning for physical systems.
method Bayesian Neural Networks (BNN), Concrete Dropout (CD), Deep Ensembles (DE), and Analytic Error Propagation.
result Pitfalls in using UQ methods, especially Bayesian Neural Networks and Concrete Dropout.
Study shows conventional data prep fails for insurance data, proposing new methods.
problem Challenges in data preparation for insurance data lead to unreliable models.
method Proposes a new data preparation framework using support points and Chatterjee correlation coefficient.
result New methods significantly enhance model robustness and reduce computational resource requirements.
Abstract MDPs enable strategic exploration and fast reward transfer in complex environments.
problem Challenging to learn accurate MDPs for high-dimensional states.
method Learn an abstract MDP over low-dimensional coarse states, using an abstraction function.
result Achieves superhuman performance on Pitfall! and higher reward with fewer samples.