Machine learning models predict out of distribution data before failure

PLSP (Pre-hoc Liminal Space Profiling): OOD Prediction over Detection -- An Anticipatory Approach for Machine Learning Model Reliability

Machine LearningComputer Vision and Pattern Recognition

Summary

Machine learning models can fail when they encounter data unlike anything they were trained on, called out-of-distribution (OOD) data. Instead of spotting OOD data only after the fact, the authors propose a new way that predicts when such data will appear and how the model will behave. They introduce a new measure called the CREDibility Score that works across different datasets to estimate model reliability ahead of deployment. Their tests show this approach helps understand and prepare for these challenging situations better than existing methods.

What this means in practice

Authors

Vipul Bansal, Himanshu Buckchash, Balasubramanian Raman, Deepak Dhungana

Abstract

Out-of-Distribution (OOD) data poses a significant threat to machine learning models, often leading to model failure during deployment. All existing OOD detection methods are post-hoc, relying on evaluation metrics such as accuracy and AUC-ROC during inference to indirectly assess the model's response to OOD data by measuring deviations. In contrast to existing approaches, the proposed work shifts the paradigm from OOD detection to OOD prediction by proposing a pre-hoc anticipatory framework called PLSP for OOD prediction. We make several key contributions: (a) a dataset-independent metric called the CREDibility Score (CREDS) is proposed for OOD prediction; (b) credibility curves are introduced to study the maximum credibility a model can attain; and (c) credibility heat maps (and volume under surface) are introduced to characterize pre-hoc model behavior across different datasets. This work provides a novel perspective on signal processing under distributional shifts. Experiments across multiple datasets demonstrate that the proposed metric serves as a valuable measure for improving the robustness of machine learning models toward OOD prediction.