Incremental linear discriminant analysis for evolving feature spaces in multitask pattern recognition problems.
Masayuki HisadaSeiichi OzawaKau ZhangNikola K. KasabovPublished in: Evol. Syst. (2010)
Keyphrases
- computer vision
- linear discriminant analysis
- pattern recognition problems
- multi task
- feature space
- pattern recognition
- dimensionality reduction
- feature selection
- discriminant analysis
- multi task learning
- face recognition
- multitask learning
- principal component analysis
- learning tasks
- dimension reduction
- support vector machine svm
- high dimensional
- feature extraction
- metric learning
- learning problems
- kernel methods
- multi class
- null space
- training samples
- high dimensional data
- high dimensionality
- machine learning
- transfer learning
- feature vectors
- low dimensional
- classification accuracy
- gaussian processes
- support vector machine
- input data
- image processing
- hyperplane
- information gain
- neural network
- unsupervised learning
- maximum margin
- feature set
- support vector
- high dimensional feature space
- multiple features
- qr decomposition
- scatter matrix
- learning algorithm
- singular value decomposition
- principal components
- data points