Entropy Reweighted Conformal Classification
This addresses an incremental improvement in conformal classification methods for researchers and practitioners needing reliable prediction sets.
The paper tackled the problem of efficiency degradation when integrating confidence calibration with Conformal Prediction for classification, and the result was a method that significantly improves efficiency through entropy-based reweighting.
Conformal Prediction (CP) is a powerful framework for constructing prediction sets with guaranteed coverage. However, recent studies have shown that integrating confidence calibration with CP can lead to a degradation in efficiency. In this paper, We propose an adaptive approach that considers the classifier's uncertainty and employs entropy-based reweighting to enhance the efficiency of prediction sets for conformal classification. Our experimental results demonstrate that this method significantly improves efficiency.