Contrastive regularization guided label refurbishment for fault diagnosis under label noise.
Jiankang ZhongYongjun YangHanling MaoAisong QinXinxin LiWeili TangPublished in: Adv. Eng. Informatics (2024)
Keyphrases
- label noise
- fault diagnosis
- active learning
- expert systems
- neural network
- noise tolerant
- fault detection
- training data
- noise model
- fuzzy logic
- electronic equipment
- fault detection and diagnosis
- bp neural network
- operating conditions
- rotating machinery
- power transformers
- condition monitoring
- multiple faults
- learning algorithm
- monitoring and fault diagnosis
- analog circuits
- chemical process
- multi sensor information fusion
- real time
- multiple instance learning
- power plant
- gas turbine
- maximum likelihood
- training set