Feature dimension reduction using reduced-rank maximum likelihood estimation for hidden Markov models.
Don X. SunPublished in: ICSLP (1996)
Keyphrases
- hidden markov models
- dimension reduction
- maximum likelihood estimation
- minimum classification error
- maximum likelihood
- em algorithm
- principal component analysis
- continuous hidden markov models
- feature extraction
- low dimensional
- high dimensional
- high dimensional data
- expectation maximization
- parameter estimation
- linear discriminant analysis
- gesture recognition
- dimensionality reduction
- singular value decomposition
- probability distribution
- feature selection
- unsupervised learning
- conditional random fields
- cluster analysis
- high dimensionality
- feature space
- density function
- discriminative training
- optical flow
- data sets
- classification error
- pattern recognition
- preprocessing
- feature vectors