# A Hierarchy of Limitations in Machine Learning

arXiv: https://arxiv.org/abs/2002.05193 · HTML: https://mlsanity.com/paper/2002.05193/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Feb 2020.

## Built on by (most cited first, 20 of 20)

- [AI and the Everything in the Whole Wide World Benchmark](https://mlsanity.com/paper/2111.15366/) (2021)
- [Leakage and the Reproducibility Crisis in ML-based Science](https://mlsanity.com/paper/2207.07048/) (2022)
- [How to avoid machine learning pitfalls: a guide for academic researchers](https://mlsanity.com/paper/2108.02497/) (2021)
- [Rethinking Fairness: An Interdisciplinary Survey of Critiques of Hegemonic ML Fairness Approaches](https://mlsanity.com/paper/2205.04460/) (2022)
- [Scalable Qualitative Coding with LLMs: Chain-of-Thought Reasoning Matches Human Performance in Some Hermeneutic Tasks](https://mlsanity.com/paper/2401.15170/) (2024)
- [The worst of both worlds: A comparative analysis of errors in learning from data in psychology and machine learning](https://mlsanity.com/paper/2203.06498/) (2022)
- [REFORMS: Reporting Standards for Machine Learning Based Science](https://mlsanity.com/paper/2308.07832/) (2023)
- [Rigor in AI: Doing Rigorous AI Work Requires a Broader, Responsible AI-Informed Conception of Rigor](https://mlsanity.com/paper/2506.14652/) (2025)
- [Limitations of machine learning for building energy prediction: ASHRAE Great Energy Predictor III Kaggle competition error analysis](https://mlsanity.com/paper/2106.13475/) (2021)
- [Mapping the Internet: Modelling Entity Interactions in Complex Heterogeneous Networks](https://mlsanity.com/paper/2104.09650/) (2021)
- [System Safety Engineering for Social and Ethical ML Risks: A Case Study](https://mlsanity.com/paper/2211.04602/) (2022)
- [Concrete Safety for ML Problems: System Safety for ML Development and Assessment](https://mlsanity.com/paper/2302.02972/) (2023)
- [Tackling Cyberattacks through AI-based Reactive Systems: A Holistic Review and Future Vision](https://mlsanity.com/paper/2312.06229/) (2023)
- [Specification Overfitting in Artificial Intelligence](https://mlsanity.com/paper/2403.08425/) (2024)
- [Talking About the Assumption in the Room](https://mlsanity.com/paper/2502.13268/) (2025)
- [Exposing Assumptions in AI Benchmarks through Cognitive Modelling](https://mlsanity.com/paper/2409.16849/) (2024)
- [Combining Self-attention and Dilation Convolutional for Semantic Segmentation of Coal Maceral Groups](https://mlsanity.com/paper/2506.12712/) (2025)
- [Adaptive Physics-Informed Neural Networks with Multi-Category Feature Engineering for Hydrogen Sorption Prediction in Clays, Shales, and Coals](https://mlsanity.com/paper/2509.00049/) (2025)
- [Every Character Counts: From Vulnerability to Defense in Phishing Detection](https://mlsanity.com/paper/2509.20589/) (2025)
- [Physics-Informed Neural Networks for Predicting Hydrogen Sorption in Geological Formations: Thermodynamically Constrained Deep Learning Integrating Classical Adsorption Theory](https://mlsanity.com/paper/2603.28328/) (2026)

## Abstract (arXiv)

"All models are wrong, but some are useful", wrote George E. P. Box (1979). Machine learning has focused on the usefulness of probability models for prediction in social systems, but is only now coming to grips with the ways in which these models are wrong---and the consequences of those shortcomings. This paper attempts a comprehensive, structured overview of the specific conceptual, procedural, and statistical limitations of models in machine learning when applied to society. Machine learning modelers themselves can use the described hierarchy to identify possible failure points and think through how to address them, and consumers of machine learning models can know what to question when confronted with the decision about if, where, and how to apply machine learning. The limitations go from commitments inherent in quantification itself, through to showing how unmodeled dependencies can lead to cross-validation being overly optimistic as a way of assessing model performance.