A flexible algorithm for fault diagnosis in a centrifugal pump with corrupted data and noise based on ANN and support vector machine with hyper-parameters optimization.
Ali AzadehMorteza SaberiAhmad KazemVahid EbrahimipourA. NourmohammadzadehZahra SaberiPublished in: Appl. Soft Comput. (2013)
Keyphrases
- fault diagnosis
- input data
- data sets
- prior information
- hyperparameters
- training data
- learning algorithm
- neural network
- missing data
- condition monitoring
- bayesian framework
- model selection
- data points
- support vector machine
- incomplete data
- prior knowledge
- cross validation
- support vector
- closed form
- expert systems
- worst case
- expectation maximization
- k means
- machine learning
- artificial intelligence
- random sampling
- noise level
- parameter settings
- incremental learning
- feature selection
- support vector machine svm
- missing values
- em algorithm
- feature vectors
- energy function
- artificial neural networks
- subband