# Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shift

arXiv: https://arxiv.org/abs/1810.11953 · HTML: https://mlsanity.com/paper/1810.11953/

## Key facts

- Cited by 243 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Oct 2018.

## Built on by (most cited first, 30 of 243)

- [Towards Causal Representation Learning](https://mlsanity.com/paper/2102.11107/) (2021)
- [Trustworthy AI: From Principles to Practices](https://mlsanity.com/paper/2110.01167/) (2021)
- [MultiBench: Multiscale Benchmarks for Multimodal Representation Learning](https://mlsanity.com/paper/2107.07502/) (2021)
- [Leveraging Unlabeled Data to Predict Out-of-Distribution Performance](https://mlsanity.com/paper/2201.04234/) (2022)
- [Towards Robust and Reliable Algorithmic Recourse](https://mlsanity.com/paper/2102.13620/) (2021)
- [Distilling Model Failures as Directions in Latent Space](https://mlsanity.com/paper/2206.14754/) (2022)
- [MMD Aggregated Two-Sample Test](https://mlsanity.com/paper/2110.15073/) (2021)
- [Mixture Proportion Estimation and PU Learning: A Modern Approach](https://mlsanity.com/paper/2111.00980/) (2021)
- [GitTables: A Large-Scale Corpus of Relational Tables](https://mlsanity.com/paper/2106.07258/) (2021)
- [Diagnosing failures of fairness transfer across distribution shift in real-world medical settings](https://mlsanity.com/paper/2202.01034/) (2022)
- [Neural Mean Discrepancy for Efficient Out-of-Distribution Detection](https://mlsanity.com/paper/2104.11408/) (2021)
- [Look Before You Leap: An Exploratory Study of Uncertainty Measurement for Large Language Models](https://mlsanity.com/paper/2307.10236/) (2023)
- [Tracking the risk of a deployed model and detecting harmful distribution shifts](https://mlsanity.com/paper/2110.06177/) (2021)
- [A survey on learning from imbalanced data streams: taxonomy, challenges, empirical study, and reproducible experimental framework](https://mlsanity.com/paper/2204.03719/) (2022)
- [How I failed machine learning in medical imaging -- shortcomings and recommendations](https://mlsanity.com/paper/2103.10292/) (2021)
- [Dataset Inference for Self-Supervised Models](https://mlsanity.com/paper/2209.09024/) (2022)
- [Calibrated Selective Classification](https://mlsanity.com/paper/2208.12084/) (2022)
- [A Validity Perspective on Evaluating the Justified Use of Data-driven Decision-making Algorithms](https://mlsanity.com/paper/2206.14983/) (2022)
- [Conditional Generative Models for Counterfactual Explanations](https://mlsanity.com/paper/2101.10123/) (2021)
- [Towards Explaining Distribution Shifts](https://mlsanity.com/paper/2210.10275/) (2022)
- [Operationalizing Machine Learning: An Interview Study](https://mlsanity.com/paper/2209.09125/) (2022)
- [Calibration of Time-Series Forecasting: Detecting and Adapting Context-Driven Distribution Shift](https://mlsanity.com/paper/2310.14838/) (2023)
- [Feature Shift Detection: Localizing Which Features Have Shifted via Conditional Distribution Tests](https://mlsanity.com/paper/2107.06929/) (2021)
- [Linking Neural Collapse and L2 Normalization with Improved Out-of-Distribution Detection in Deep Neural Networks](https://mlsanity.com/paper/2209.08378/) (2022)
- [A Geometric Explanation of the Likelihood OOD Detection Paradox](https://mlsanity.com/paper/2403.18910/) (2024)
- [AutoML Two-Sample Test](https://mlsanity.com/paper/2206.08843/) (2022)
- [Deconfounded Representation Similarity for Comparison of Neural Networks](https://mlsanity.com/paper/2202.00095/) (2022)
- [Estimating and Explaining Model Performance When Both Covariates and Labels Shift](https://mlsanity.com/paper/2209.08436/) (2022)
- [A Learning Based Hypothesis Test for Harmful Covariate Shift](https://mlsanity.com/paper/2212.02742/) (2022)
- ["Why did the Model Fail?": Attributing Model Performance Changes to Distribution Shifts](https://mlsanity.com/paper/2210.10769/) (2022)

## Abstract (arXiv)

We might hope that when faced with unexpected inputs, well-designed software systems would fire off warnings. Machine learning (ML) systems, however, which depend strongly on properties of their inputs (e.g. the i.i.d. assumption), tend to fail silently. This paper explores the problem of building ML systems that fail loudly, investigating methods for detecting dataset shift, identifying exemplars that most typify the shift, and quantifying shift malignancy. We focus on several datasets and various perturbations to both covariates and label distributions with varying magnitudes and fractions of data affected. Interestingly, we show that across the dataset shifts that we explore, a two-sample-testing-based approach, using pre-trained classifiers for dimensionality reduction, performs best. Moreover, we demonstrate that domain-discriminating approaches tend to be helpful for characterizing shifts qualitatively and determining if they are harmful.