Deep continual transfer learning with dynamic weight aggregation for fault diagnosis of industrial streaming data under varying working conditions.
Jipu LiRuyi HuangZhuyun ChenGuolin HeKonstantinos C. GrylliasWeihua LiPublished in: Adv. Eng. Informatics (2023)
Keyphrases
- fault diagnosis
- transfer learning
- streaming data
- industrial systems
- gas turbine
- chemical process
- operating conditions
- data streams
- neural network
- fault detection
- sliding window
- expert systems
- knowledge transfer
- labeled data
- reinforcement learning
- concept drift
- transfer knowledge
- text classification
- fuzzy logic
- active learning
- collaborative filtering
- industrial processes
- data distribution
- fault diagnostic
- industrial applications
- machine learning
- semi supervised learning
- class labels
- text categorization
- previously learned
- multi dimensional
- text mining
- semi supervised
- information extraction
- learning algorithm