FDI-X: An Occlusion-based Approach for Improving the Explainability of Deep Learning Models in Fault Detection and Isolation.
Balyogi Mohan DashBelkacem Ould BouamamaKomi Midzodzi PekpeMahdi BoukerdjaPublished in: ICCAD (2023)
Keyphrases
- learning models
- fault detection and isolation
- industrial processes
- machine learning
- learning tasks
- fault detection
- learning algorithm
- fault diagnosis
- loss function
- machine learning algorithms
- semi supervised learning
- closed loop
- conditional random fields
- learning problems
- classification models
- control system
- bayesian networks
- data sets
- decision trees
- decision making
- knowledge base
- fuzzy logic
- training data