Augmenting Monte Carlo Dropout Classification Models with Unsupervised Learning Tasks for Detecting and Diagnosing Out-of-Distribution Faults

TL;DR

This paper enhances Monte Carlo dropout classifiers by integrating unsupervised learning tasks, significantly improving out-of-distribution fault detection and diagnosis across multiple datasets.

Contribution

It introduces a novel augmentation of dropout-based models with unsupervised tasks, justified through information theory, to better detect incipient and unknown faults.

Findings

Improved fault detection accuracy on diverse datasets.

Enhanced diagnosis of out-of-distribution faults.

Effective integration of unsupervised learning with uncertainty estimation.

Abstract

The Monte Carlo dropout method has proved to be a scalable and easy-to-use approach for estimating the uncertainty of deep neural network predictions. This approach was recently applied to Fault Detection and Di-agnosis (FDD) applications to improve the classification performance on incipient faults. In this paper, we propose a novel approach of augmenting the classification model with an additional unsupervised learning task. We justify our choice of algorithm design via an information-theoretical analysis. Our experimental results on three datasets from diverse application domains show that the proposed method leads to improved fault detection and diagnosis performance, especially on out-of-distribution examples including both incipient and unknown faults.