Data-driven Kernel-based Probabilistic SAX for Time Series Dimensionality Reduction.
Konstantinos BountrogiannisGeorge TzagkarakisPanagiotis TsakalidesPublished in: EUSIPCO (2020)
Keyphrases
- data driven
- dimensionality reduction
- kernel pca
- high dimensional
- principal component analysis
- symbolic aggregate approximation
- feature extraction
- kernel discriminant analysis
- dynamic time warping
- data representation
- pattern recognition
- low dimensional
- high dimensional data
- principal components
- random projections
- kernel trick
- high dimensionality
- discriminant analysis
- manifold learning
- probabilistic model
- feature space
- symbolic representation
- linear dimensionality reduction
- structure preserving
- multivariate time series
- dimensionality reduction methods
- support vector
- linear discriminant analysis
- kernel methods
- support vector machine
- data sets
- discrete valued
- subsequence matching
- metric learning
- uncertain data
- euclidean distance
- generative model