Data Heterogeneity Modeling for Trustworthy Machine Learning
Not on the map. Why?
Abstract (arXiv)
Data heterogeneity plays a pivotal role in determining the performance of machine learning (ML) systems. Traditional algorithms, which are typically designed to optimize average performance, often overlook the intrinsic diversity within datasets. This oversight can lead to a myriad of issues, including unreliable decision-making, inadequate generalization across different domains, unfair outcomes, and false scientific inferences. Hence, a nuanced approach to modeling data heterogeneity is essential for the development of dependable, data-driven systems. In this survey paper, we present a thorough exploration of heterogeneity-aware machine learning, a paradigm that systematically integrates considerations of data heterogeneity throughout the entire ML pipeline -- from data collection and model training to model evaluation and deployment. By applying this approach to a variety of critical fields, including healthcare, agriculture, finance, and recommendation systems, we demonstrate the substantial benefits and potential of heterogeneity-aware ML. These applications underscore how a deeper understanding of data diversity can enhance model robustness, fairness, and reliability and help model diagnosis and improvements. Moreover, we delve into future directions and provide research opportunities for the whole data mining community, aiming to promote the development of heterogeneity-aware ML.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 57 of its 93 references that are in the map. First posted on arXiv 1 Jun 2025.
Builds on
57 of 93 references.
References · 57
- 2016Equality of Opportunity in Supervised Learningcited by 2,007
- 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,917
- 2019Invariant Risk Minimizationcited by 1,854
- 2017Understanding Black-box Predictions via Influence Functionscited by 1,741
- 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,375
- 2019Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalizationcited by 1,247
- 2020In Search of Lost Domain Generalizationcited by 1,054
- 2017Deeper, Broader and Artier Domain Generalizationcited by 986
- 2021Domain Generalization: A Surveycited by 764
- 2018An Empirical Study of Example Forgetting during Deep Neural Network Learningcited by 678
- 2015Estimation and Inference of Heterogeneous Treatment Effects using Random Forestscited by 600
- 2021Just Train Twice: Improving Group Robustness without Training Group Informationcited by 523
38 more
- 2021Retiring Adult: New Datasets for Fair Machine Learningcited by 449
- 2021Towards Out-Of-Distribution Generalization: A Surveycited by 436
- 2020Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamicscited by 425
- 2018Learning Models with Uniform Performance via Distributionally Robust Optimizationcited by 335
- 2023Studying Large Language Model Generalization with Influence Functionscited by 326
- 2016Generalized Random Forestscited by 312
- 2020Environment Inference for Invariant Learningcited by 306
- 2015Data-driven Distributionally Robust Optimization Using the Wasserstein Metric: Performance Guarantees and Tractable Reformulationscited by 291
- 2020Identifying Mislabeled Data using the Area Under the Margin Rankingcited by 261
- 2021Accuracy on the Line: On the Strong Correlation Between Out-of-Distribution and In-Distribution Generalizationcited by 249
- 2021Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalizationcited by 220
- 2020Invariant Risk Minimization Gamescited by 188
- 2022Handling Distribution Shifts on Graphs: An Invariance Perspectivecited by 164
- 2020Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scanscited by 158
- 2019Data Valuation using Reinforcement Learningcited by 116
- 2021Heterogeneous Risk Minimizationcited by 90
- 2022NICO++: Towards Better Benchmarking for Domain Generalizationcited by 86
- 2019Distributionally Robust Optimization and Generalization in Kernel Methodscited by 73
- 2019Stable Learning via Sample Reweightingcited by 70
- 2020Distributionally Robust Losses for Latent Covariate Mixturescited by 58
- 2013Challenges of Big Data Analysiscited by 47
- 2021On Tilted Losses in Machine Learning: Theory and Applicationscited by 43
- 2023Does Invariant Graph Learning via Environment Augmentation Learn Invariance?cited by 42
- 2022Turning the Tables: Biased, Imbalanced, Dynamic Tabular Datasets for ML Evaluationcited by 40
- 2024Top Leaderboard Ranking = Top Coding Proficiency, Always? EvoEval: Evolving Coding Benchmarks via LLMcited by 40
- 2023Diagnosing Model Performance Under Distribution Shiftcited by 39
- 2024Distributionally Robust Optimization and Robust Statisticscited by 24
- 2022Data-IQ: Characterizing subgroups with heterogeneous outcomes in tabular datacited by 22
- 2021Kernelized Heterogeneous Risk Minimizationcited by 21
- 2022Subgroup Robustness Grows On Trees: An Empirical Baseline Investigationcited by 17
- 2020Stable Adversarial Learning under Distributional Shiftscited by 16
- 2017Data-driven Optimal Cost Selection for Distributionally Robust Optimizationcited by 15
- 2021Distributionally Robust Learning with Stable Adversarial Trainingcited by 14
- 2022Evaluating Robustness to Dataset Shift via Parametric Robustness Setscited by 10
- 2023Where Does My Model Underperform? A Human Evaluation of Slice Discovery Algorithmscited by 9
- 2023TRIAGE: Characterizing and auditing training data for improved regressioncited by 6
- 2023Geometry-Calibrated DRO: Combating Over-Pessimism with Free Energy Implicationscited by 5
- 2024A hierarchical decomposition for explaining ML performance discrepanciescited by 5
50 most cited of 57 All 57 (JSON)
Built on by
Not cited yet.