A Partial-Label U-Net Learning Method for Compound-Fault Diagnosis With Fault- Sample Class Imbalance.
Jingfei ZhangXiao HePublished in: IEEE Trans. Ind. Informatics (2024)
Keyphrases
- fault diagnosis
- multi sensor information fusion
- prior knowledge
- fault tree
- bp neural network
- expert systems
- pairwise
- input data
- unsupervised learning
- class imbalance
- multiple faults
- fault detection
- semi supervised learning
- fault detection and isolation
- operating conditions
- error rate
- artificial intelligence
- active learning
- reinforcement learning
- decision making
- feature selection