Saved in:
Bibliographic Details
Main Authors: Wang, Cheng, Golebiowski, Jacek
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2303.15057
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911931198078976
author Wang, Cheng
Golebiowski, Jacek
author_facet Wang, Cheng
Golebiowski, Jacek
contents Model miscalibration has been frequently identified in modern deep neural networks. Recent work aims to improve model calibration directly through a differentiable calibration proxy. However, the calibration produced is often biased due to the binning mechanism. In this work, we propose to learn better-calibrated models via meta-regularization, which has two components: (1) gamma network (gamma-net), a meta learner that outputs sample-wise gamma values (continuous variable) for Focal loss for regularizing the backbone network; (2) smooth expected calibration error (SECE), a Gaussian-kernel based, unbiased, and differentiable surrogate to ECE that enables the smooth optimization of gamma-Net. We evaluate the effectiveness of the proposed approach in regularizing neural networks towards improved and unbiased calibration on three computer vision datasets. We empirically demonstrate that: (a) learning sample-wise gamma as continuous variables can effectively improve calibration; (b) SECE smoothly optimizes gamma-net towards unbiased and robust calibration with respect to the binning schemes; and (c) the combination of gamma-net and SECE achieves the best calibration performance across various calibration metrics while retaining very competitive predictive performance as compared to multiple recently proposed methods.
format Preprint
id arxiv_https___arxiv_org_abs_2303_15057
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Towards Unbiased Calibration using Meta-Regularization
Wang, Cheng
Golebiowski, Jacek
Machine Learning
Model miscalibration has been frequently identified in modern deep neural networks. Recent work aims to improve model calibration directly through a differentiable calibration proxy. However, the calibration produced is often biased due to the binning mechanism. In this work, we propose to learn better-calibrated models via meta-regularization, which has two components: (1) gamma network (gamma-net), a meta learner that outputs sample-wise gamma values (continuous variable) for Focal loss for regularizing the backbone network; (2) smooth expected calibration error (SECE), a Gaussian-kernel based, unbiased, and differentiable surrogate to ECE that enables the smooth optimization of gamma-Net. We evaluate the effectiveness of the proposed approach in regularizing neural networks towards improved and unbiased calibration on three computer vision datasets. We empirically demonstrate that: (a) learning sample-wise gamma as continuous variables can effectively improve calibration; (b) SECE smoothly optimizes gamma-net towards unbiased and robust calibration with respect to the binning schemes; and (c) the combination of gamma-net and SECE achieves the best calibration performance across various calibration metrics while retaining very competitive predictive performance as compared to multiple recently proposed methods.
title Towards Unbiased Calibration using Meta-Regularization
topic Machine Learning
url https://arxiv.org/abs/2303.15057