Missing-Data Classification With the Extended Full-Dimensional Gaussian Mixture Model: Applications to EMG-Based Motion Recognition.
Qichuan DingJianda HanXingang ZhaoYang ChenPublished in: IEEE Trans. Ind. Electron. (2015)
Keyphrases
- missing data
- gaussian mixture model
- finite mixture models
- feature vectors
- multiple imputation
- motion recognition
- em algorithm
- missing values
- incomplete data
- feature space
- mixture model
- motion segmentation
- expectation maximization
- feature extraction
- maximum likelihood criterion
- background subtraction
- maximum likelihood
- pattern recognition
- text classification
- support vector machine
- model selection
- motion estimation
- human motion
- human actions
- unsupervised learning
- space time
- generative model
- pairwise
- similarity measure