mlsanityData to 14 Oct 2026

Sources of Irreproducibility in Machine Learning: A Review

arXiv 2204.07610 ↗First posted 15 Apr 2022cs.LG cs.AIcited by 32

Not on the map. Why?

Abstract (arXiv)

Background: Many published machine learning studies are irreproducible. Issues with methodology and not properly accounting for variation introduced by the algorithm themselves or their implementations are attributed as the main contributors to the irreproducibility.Problem: There exist no theoretical framework that relates experiment design choices to potential effects on the conclusions. Without such a framework, it is much harder for practitioners and researchers to evaluate experiment results and describe the limitations of experiments. The lack of such a framework also makes it harder for independent researchers to systematically attribute the causes of failed reproducibility experiments. Objective: The objective of this paper is to develop a framework that enable applied data science practitioners and researchers to understand which experiment design choices can lead to false findings and how and by this help in analyzing the conclusions of reproducibility experiments. Method: We have compiled an extensive list of factors reported in the literature that can lead to machine learning studies being irreproducible. These factors are organized and categorized in a reproducibility framework motivated by the stages of the scientific method. The factors are analyzed for how they can affect the conclusions drawn from experiments. A model comparison study is used as an example. Conclusion: We provide a framework that describes machine learning methodology from experimental design decisions to the conclusions inferred from them.

Key facts. Cited by 32 papers in the map (data through 14 Oct 2026). Builds on 25 of its 103 references that are in the map. First posted on arXiv 15 Apr 2022.

Builds on

25 of 103 references.

References · 25

  1. 2019Do ImageNet Classifiers Generalize to ImageNet?cited by 1,515
  2. 2018Datasheets for Datasetscited by 1,386
  3. 2017Deep Reinforcement Learning that Matterscited by 754
  4. 2020Improving Reproducibility in Machine Learning Research (A Report from the NeurIPS 2019 Reproducibility Program)cited by 322
  5. 2017Are GANs Created Equal? A Large-Scale Studycited by 267
  6. 2021Accounting for Variance in Machine Learning Benchmarkscited by 201
  7. 2019Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Recommendation Approachescited by 185
  8. 2019Show Your Work: Improved Reporting of Experimental Resultscited by 169
  9. 2020Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scanscited by 158
  10. 2020With Little Power Comes Great Responsibilitycited by 142
  11. 2018Troubling Trends in Machine Learning Scholarshipcited by 108
  12. 2017On the State of the Art of Evaluation in Neural Language Modelscited by 104
13 more
  1. 2019A Step Toward Quantifying Independently Reproducible Machine Learning Researchcited by 77
  2. 2017Reporting Score Distributions Makes a Difference: Performance Study of LSTM-networks for Sequence Taggingcited by 73
  3. 2019A Troubling Analysis of Reproducibility and Progress in Recommender Systems Researchcited by 65
  4. 2021Randomness In Neural Network Training: Characterizing The Impact of Toolingcited by 59
  5. 2021A Systematic Review of Reproducibility Research in Natural Language Processingcited by 34
  6. 2018A Large-Scale Study on Regularization and Normalization in GANscited by 26
  7. 2021Hyperparameter Optimization Is Deceiving Us, and How to Stop Itcited by 20
  8. 2020The Fundamental Principles of Reproducibilitycited by 17
  9. 2022The worst of both worlds: A comparative analysis of errors in learning from data in psychology and machine learningcited by 17
  10. 2020Research Reproducibility as a Survival Analysiscited by 16
  11. 2020Session-aware Recommendation: A Surprising Quest for the State-of-the-artcited by 15
  12. 2020Critical analysis on the reproducibility of visual quality assessment using deep featurescited by 5
  13. 2021On the reproducibility of fully convolutional neural networks for modeling time-space evolving physical systemscited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 2; 2023: 7; 2024: 10; 2025: 6; 2026: 7.

32 papers.

Sort:

By area:

  1. 2024Weak baselines and reporting biases lead to overoptimism in machine learning for fluid-related partial differential equationscited by 91
  2. 2022Efficient Methods for Natural Language Processing: A Surveycited by 78
  3. 2024Reproducibility in Machine Learning-based Research: Overview, Barriers and Driverscited by 34
  4. 2024Benchmark Data Repositories for Better Benchmarkingcited by 10
  5. 2024EBES: Easy Benchmarking for Event Sequencescited by 7
  6. 2023A Two-Sided Discussion of Preregistration of NLP Researchcited by 6
  7. 2023Towards Inferential Reproducibility of Machine Learning Researchcited by 4
  8. 2025The Impact of Feature Scaling In Machine Learning: Effects on Regression and Classification Taskscited by 4
  9. 2023Examining the Effect of Implementation Factors on Deep Learning Reproducibilitycited by 3
  10. 2024Towards a Categorical Foundation of Deep Learning: A Surveycited by 2
  11. 2025AI Copilots for Reproducibility in Science: A Case Studycited by 2
  12. 2026How Open Must Language Models be to Enable Reliable Scientific Inference?cited by 2
20 more
  1. 2022Online learning techniques for prediction of temporal tabular datasets with regime changescited by 1
  2. 2023A Rigorous Uncertainty-Aware Quantification Framework Is Essential for Reproducible and Replicable Machine Learning Workflowscited by 1
  3. 2023Deep incremental learning models for financial temporal tabular datasets with distribution shiftscited by 1
  4. 2023Robust Nonparametric Hypothesis Testing to Understand Variability in Training Neural Networkscited by 1
  5. 2024Measuring training variability from stochastic optimization using robust nonparametric testingcited by 1
  6. 2024The Unreasonable Effectiveness of Open Science in AI: A Replication Studycited by 1
  7. 2025Statistical Quality and Reproducibility of Pseudorandom Number Generators in Machine Learning technologiescited by 1
  8. 2026The Shift Toward Open and Reproducible AI Researchcited by 1
  9. 2023An End-to-End System for Reproducibility Assessment of Source Code Repositories via Their Readmescited by 0
  10. 2024A Second Look on BASS -- Boosting Abstractive Summarization with Unified Semantic Graphs -- A Replication Studycited by 0
  11. 2024Reproducibility and Geometric Intrinsic Dimensionality: An Investigation on Graph Neural Network Researchcited by 0
  12. 2024Generalizability of experimental studiescited by 0
  13. 2025Aerial Image Classification in Scarce and Unconstrained Environments via Conformal Predictioncited by 0
  14. 2025Reducing Variability of Multiple Instance Learning Methods for Digital Pathologycited by 0
  15. 2025Fractal basin geometry and the limits to predictability and reproducibility in deep learningcited by 0
  16. 2026Position: Ideas Should be the Center of Machine Learning Researchcited by 0
  17. 2026I-CARE: Analysis of interference-related phenomena in a controllable, diverse and representative unlearning setting for text-to-image modelscited by 0
  18. 2026Measuring consistency via ensemble margin and local prediction variability: Auditing decision systems in the presence of predictive multiplicitycited by 0
  19. 2026RECLAIM: Can Agents Reproduce the Claims of Machine Learning Papers?cited by 0
  20. 2026Kernel-based External Validity of Experimental Studiescited by 0