Estimating Uncertainty with Implicit Quantile Network

2024/08/26 by Yi Hung Lim, Lim, Yi Hung · 1 citation
Engineering · Computer Science · #Fault Detection and Control Systems #Neural Networks and Applications

paper · pdf · doi:10.48550/arxiv.2408.14525

Abstract

Uncertainty quantification is an important part of many performance critical applications. This paper provides a simple alternative to existing approaches such as ensemble learning and bayesian neural networks. By directly modeling the loss distribution with an Implicit Quantile Network, we get an estimate of how uncertain the model is of its predictions. For experiments with MNIST and CIFAR datasets, the mean of the estimated loss distribution is 2x higher for incorrect predictions. When data with high estimated uncertainty is removed from the test dataset, the accuracy of the model goes up as much as 10%. This method is simple to implement while offering important information to applications where the user has to know when the model could be wrong (e.g. deep learning for healthcare).

Cited by

• Foundation Models for Clean Energy Forecasting: A Comprehensive Review

Related

• QuantProb: Generalizing Probabilities along with Predictions for a Pre-trained Classifier
• Uncertainty Estimation Using a Single Deep Deterministic Neural Network
• Uncertainty Quantification with the Empirical Neural Tangent Kernel
• Exploring Uncertainty in Deep Learning for Construction of Prediction Intervals
• Novel Uncertainty Framework for Deep Learning Ensembles