# Automatically detecting data drift in machine learning classifiers

arXiv: https://arxiv.org/abs/2111.05672 · HTML: https://mlsanity.com/paper/2111.05672/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- Builds on 1 of its 24 references that are in the map.
- First posted on arXiv 10 Nov 2021.

## Builds on (most cited first, 1 of 1)

- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)

## Built on by (most cited first, 15 of 15)

- [PRSA: Prompt Stealing Attacks against Real-World Prompt Services](https://mlsanity.com/paper/2402.19200/) (2024)
- [CheXstray: Real-time Multi-Modal Data Concordance for Drift Detection in Medical Imaging AI](https://mlsanity.com/paper/2202.02833/) (2022)
- [Test & Evaluation Best Practices for Machine Learning-Enabled Systems](https://mlsanity.com/paper/2310.06800/) (2023)
- [Systematic Literature Review on Application of Machine Learning in Continuous Integration](https://mlsanity.com/paper/2305.12695/) (2023)
- [Reliable and Interpretable Drift Detection in Streams of Short Texts](https://mlsanity.com/paper/2305.17750/) (2023)
- [Towards the Visualization of Aggregated Class Activation Maps to Analyse the Global Contribution of Class Features](https://mlsanity.com/paper/2308.00710/) (2023)
- [Frouros: A Python library for drift detection in machine learning systems](https://mlsanity.com/paper/2208.06868/) (2022)
- [Data Quality Aware Approaches for Addressing Model Drift of Semantic Segmentation Models](https://mlsanity.com/paper/2402.07258/) (2024)
- [Model Monitoring: A General Framework with an Application to Non-life Insurance Pricing](https://mlsanity.com/paper/2510.04556/) (2025)
- [A Framework for Monitoring and Retraining Language Models in Real-World Applications](https://mlsanity.com/paper/2311.09930/) (2023)
- [Trimming the Risk: Towards Reliable Continuous Training for Deep Learning Inspection Systems](https://mlsanity.com/paper/2409.09108/) (2024)
- [Privacy Drift: Evolving Privacy Concerns in Incremental Learning](https://mlsanity.com/paper/2412.05183/) (2024)
- [Enhancing Deployment-Time Predictive Model Robustness for Code Analysis and Optimization](https://mlsanity.com/paper/2501.00298/) (2024)
- [Staying Alive: Online Neural Network Maintenance and Systemic Drift](https://mlsanity.com/paper/2503.17681/) (2025)
- [Drift Happens: An Empirical Study of Neural Architecture Robustness to Temporal Distribution Shift](https://mlsanity.com/paper/2607.05908/) (2026)

## Abstract (arXiv)

Classifiers and other statistics-based machine learning (ML) techniques generalize, or learn, based on various statistical properties of the training data. The assumption underlying statistical ML resulting in theoretical or empirical performance guarantees is that the distribution of the training data is representative of the production data distribution. This assumption often breaks; for instance, statistical distributions of the data may change. We term changes that affect ML performance `data drift' or `drift'. Many classification techniques compute a measure of confidence in their results. This measure might not reflect the actual ML performance. A famous example is the Panda picture that is correctly classified as such with a confidence of about 60%, but when noise is added it is incorrectly classified as a Gibbon with a confidence of above 99%. However, the work we report on here suggests that a classifier's measure of confidence can be used for the purpose of detecting data drift. We propose an approach based solely on classifier suggested labels and its confidence in them, for alerting on data distribution or feature space changes that are likely to cause data drift. Our approach identities degradation in model performance and does not require labeling of data in production which is often lacking or delayed. Our experiments with three different data sets and classifiers demonstrate the effectiveness of this approach in detecting data drift. This is especially encouraging as the classification itself may or may not be correct and no model input data is required. We further explore the statistical approach of sequential change-point tests to automatically determine the amount of data needed in order to identify drift while controlling the false positive rate (Type-1 error).