A Hybrid System with Hidden Markov Models and Gaussian Mixture Models for Myocardial Infarction Classification with 12-Lead ECGs.
Pei-Chann ChangJui-Chien HsiehJyun-Jie LinYen-Hung ChouChen-Hao LiuPublished in: HPCC (2009)
Keyphrases
- hidden markov models
- gaussian mixture model
- feature vectors
- sequence classification
- myocardial infarction
- mixture model
- density estimation
- speech recognition
- speaker recognition
- expectation maximization
- maximum likelihood criterion
- text classification
- feature space
- em algorithm
- pattern recognition
- support vector
- model selection
- automatic speech recognition
- feature selection
- feature extraction
- probability density function
- video sequences
- visual speech
- similarity measure
- training data
- bayesian networks
- supervised learning
- class labels
- maximum likelihood