Comparing genetic algorithms to principal component analysis and linear discriminant analysis in reducing feature dimensionality for speaker recognition.
Maider ZamalloaLuis Javier RodríguezMikel PeñagarikanoGermán BordelJuan Pedro UribePublished in: GECCO (2008)
Keyphrases
- linear discriminant analysis
- principal component analysis
- speaker recognition
- dimensionality reduction
- feature space
- genetic algorithm
- discriminant analysis
- gaussian mixture model
- dimension reduction
- feature extraction
- high dimensional data
- face recognition
- locality preserving projections
- feature vectors
- speaker identification
- vector quantization
- lower dimensional
- low dimensional
- principal components
- speaker verification
- independent component analysis
- pattern recognition
- feature set
- high dimensional
- high dimensionality
- probabilistic neural network
- face images
- unsupervised learning
- fisher criterion
- data points
- feature selection
- support vector machine svm
- classification accuracy
- neural network
- maximum likelihood
- support vector
- speech recognition
- nearest neighbor
- training set
- data analysis
- data sets