ForeCal: Random Forest-based Calibration for DNNs

TL;DR

ForeCal introduces a non-parametric, Random Forest-based post-hoc calibration method that improves the calibration of DNN classifiers while maintaining their discriminative ability, outperforming existing calibration techniques across diverse datasets.

Contribution

The paper presents ForeCal, a novel calibration algorithm leveraging Random Forests' properties to enhance calibration accuracy of DNN outputs, surpassing current methods.

Findings

ForeCal achieves lower Expected Calibration Error (ECE) than existing methods.

ForeCal maintains the AUC of the original DNN classifiers.

Experiments on 43 datasets demonstrate ForeCal's superior calibration performance.

Abstract

Deep neural network(DNN) based classifiers do extremely well in discriminating between observations, resulting in higher ROC AUC and accuracy metrics, but their outputs are often miscalibrated with respect to true event likelihoods. Post-hoc calibration algorithms are often used to calibrate the outputs of these classifiers. Methods like Isotonic regression, Platt scaling, and Temperature scaling have been shown to be effective in some cases but are limited by their parametric assumptions and/or their inability to capture complex non-linear relationships. We propose ForeCal - a novel post-hoc calibration algorithm based on Random forests. ForeCal exploits two unique properties of Random forests: the ability to enforce weak monotonicity and range-preservation. It is more powerful in achieving calibration than current state-of-the-art methods, is non-parametric, and can incorporate exogenous information as features to learn a better calibration function. Through experiments on 43 diverse datasets from the UCI ML repository, we show that ForeCal outperforms existing methods in terms of Expected Calibration Error(ECE) with minimal impact on the discriminative power of the base DNN as measured by AUC.