An Unsupervised Bayesian OC-SVM Approach for Early Degradation Detection, Thresholding, and Fault Prediction in Machinery Monitoring.
Stanley FongSriram NarasimhanPublished in: IEEE Trans. Instrum. Meas. (2022)
Keyphrases
- support vector
- support vector machine svm
- detection algorithm
- data driven
- real time
- object detection
- detection method
- unsupervised learning
- machine learning
- partially supervised
- automatic detection
- detection rate
- edge detection
- multi class
- feature selection
- support vector machine
- denoising
- monitoring system
- image processing
- svm classifier
- training set
- knn
- gray level
- false positives
- image segmentation
- post processing
- feature extraction
- maximum likelihood
- bayesian inference
- false alarms
- posterior distribution
- support vectors
- bayesian learning
- anomaly detection