Feature selection and classification algorithm for non-destructive detecting of high-speed rail defects based on vibration signals.
Mingjian SunYan WangXin ZhangYipeng LiuQiang WeiYi ShenNaizhang FengPublished in: I2MTC (2014)
Keyphrases
- classification algorithm
- high speed
- feature selection
- knn
- support vector machine
- naive bayes
- k nearest neighbor
- accurate classification
- text categorization
- condition monitoring
- vibration signal
- training set
- model selection
- class labels
- text classification
- concept drift
- feature set
- real time
- learning algorithm
- machine learning
- classification accuracy
- unsupervised learning
- support vector
- operating conditions
- feature extraction
- nearest neighbor
- active learning
- feature space
- fault diagnosis
- bayesian networks
- feature subset
- training data
- fault detection