Toward understandable semi-supervised learning fault diagnosis of chemical processes based on long short-term memory ladder autoencoder (LSTM-LAE) and self-attention (SA).
Yang JingXiaolong GeBotan LiuPublished in: Comput. Chem. Eng. (2024)
Keyphrases
- fault diagnosis
- semi supervised learning
- recurrent neural networks
- long short term memory
- neural network
- unlabeled data
- semi supervised
- labeled data
- supervised learning
- feed forward
- expert systems
- fault detection
- co training
- artificial neural networks
- semi supervised classification
- unsupervised learning
- machine learning
- training data
- multiple faults
- operating conditions
- simulated annealing
- multi sensor information fusion
- fuzzy logic
- transfer learning
- power transformers
- analog circuits
- manifold regularization
- rbf neural network
- label propagation
- labeled and unlabeled data
- pattern recognition
- active learning
- chemical process
- genetic algorithm
- learning algorithm
- artificial intelligence
- monitoring and fault diagnosis
- back propagation