PCA-KL: a parametric dimensionality reduction approach for unsupervised metric learning.
Alexandre L. M. LevadaPublished in: Adv. Data Anal. Classif. (2021)
Keyphrases
- metric learning
- dimensionality reduction
- principal component analysis
- unsupervised learning
- distance metric learning
- fully supervised
- low dimensional
- principal components analysis
- high dimensional
- principal components
- feature space
- semi supervised
- high dimensional data
- linear discriminant analysis
- dimensionality reduction methods
- feature selection
- mahalanobis distance
- pattern recognition
- linear transformation
- feature extraction
- linear dimensionality reduction
- dimension reduction
- kernel matrix
- preprocessing step
- manifold learning
- input space
- euclidean distance
- random projections
- face recognition
- discriminant analysis
- semi supervised learning
- data points
- kernel pca
- covariance matrix
- graph embedding
- pairwise
- mahalanobis metric
- kernel learning
- collaborative filtering
- locally linear embedding
- reinforcement learning