Well-calibrated Model Uncertainty with Temperature Scaling for Dropout Variational Inference

09/30/2019

∙

In this paper, well-calibrated model uncertainty is obtained by using temperature scaling together with Monte Carlo dropout as approximation to Bayesian inference. The proposed approach can easily be derived from frequentist temperature scaling and yields well-calibrated model uncertainty as well as softmax likelihood.

READ FULL TEXT

Well-calibrated Model Uncertainty with Temperature Scaling for Dropout Variational Inference

Calibration of Model Uncertainty for Dropout Variational Inference

Layer-Stack Temperature Scaling

Dropout Inference with Non-Uniform Weight Scaling

Calibration with Bias-Corrected Temperature Scaling Improves Domain Adaptation Under Label Shift in Modern Neural Networks

Distribution Calibration for Out-of-Domain Detection with Bayesian Approximation

Inference and model determination for Temperature-Driven non-linear Ecological Models

Testing whether a Learning Procedure is Calibrated

Well-calibrated Model Uncertainty with Temperature Scaling for Dropout Variational Inference

Related Research

Calibration of Model Uncertainty for Dropout Variational Inference

Layer-Stack Temperature Scaling

Dropout Inference with Non-Uniform Weight Scaling

Calibration with Bias-Corrected Temperature Scaling Improves Domain Adaptation Under Label Shift in Modern Neural Networks

Distribution Calibration for Out-of-Domain Detection with Bayesian Approximation

Inference and model determination for Temperature-Driven non-linear Ecological Models

Testing whether a Learning Procedure is Calibrated