Reliability Evaluation of Individual Predictions: A Data-centric Approach

Nima Shahbazi,Abolfazl Asudeh

2024-04-11

Abstract:Machine learning models only provide probabilistic guarantees on the expected loss of random samples from the distribution represented by their training data. As a result, a model with high accuracy, may or may not be reliable for predicting an individual query point. To address this issue, XAI aims to provide explanations of individual predictions, while approaches such as conformal predictions, probabilistic predictions, and prediction intervals count on the model's certainty in its prediction to identify unreliable cases.

Databases