Improving robustness and calibration in ensembles with diversity regularization

Abstract

Calibration and uncertainty estimation are crucial topics in high-risk environments. We introduce a new diversity regularizer for classification tasks that uses out-of-distribution samples and increases the overall accuracy, calibration and out-of-distribution detection capabilities of ensembles. Following the recent interest in the diversity of ensembles, we systematically evaluate the viability of explicitly regularizing ensemble diversity to improve calibration on in-distribution data as well as under dataset shift. We demonstrate that diversity regularization is highly beneficial in architectures, where weights are partially shared between the individual members and even allows to use fewer ensemble members to reach the same level of robustness. Experiments on CIFAR-10, CIFAR-100, and SVHN show that regularizing diversity can have a significant impact on calibration and robustness, as well as out-of-distribution detection.

References in corpus (2)

• Determinantal point processes for machine learning
• Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time

Cited by in corpus (4)

• Deep Anti-Regularized Ensembles provide reliable out-of-distribution uncertainty quantification
• Bayesian posterior approximation with stochastic ensembles
• Label-Free Model Failure Detection for Lidar-based Point Cloud Segmentation
• Deep Out-of-Distribution Uncertainty Quantification via Weight Entropy Maximization