Fully unsupervised fault detection and identification based on recursive density estimation and self-evolving cloud-based classifier.
Bruno Sielly Jales CostaPlamen Parvanov AngelovLuiz Affonso GuedesPublished in: Neurocomputing (2015)
Keyphrases
- density estimation
- fault detection
- fully unsupervised
- density estimates
- fault diagnosis
- mixture model
- industrial processes
- density function
- tennessee eastman
- probability density function
- parzen window
- probability density
- gaussian mixture model
- fuel cell
- feature selection
- training data
- feature space
- decision trees
- support vector
- training set
- power plant
- em algorithm
- svm classifier
- outlier detection
- linear classifiers
- expectation maximization
- support vector machine
- feature set
- training samples
- expert systems
- data analysis
- fuzzy logic
- feature extraction
- image segmentation
- density estimators
- neural network