Entropic Laplacian eigenmaps for unsupervised metric learning.
Alexandre Luis Magalhaes LevadaMichel Ferreira Cardia HaddadPublished in: SIBGRAPI (2021)
Keyphrases
- metric learning
- maximum variance unfolding
- laplacian eigenmaps
- semi supervised
- manifold learning
- dimensionality reduction
- kernel pca
- kernel matrix
- nonlinear dimensionality reduction
- manifold structure
- unsupervised learning
- distance metric
- pairwise
- semi supervised learning
- labeled data
- locally linear embedding
- unlabeled data
- learning tasks
- multi task
- supervised learning
- graph laplacian
- low dimensional
- feature space
- distance function
- high dimensional
- preprocessing step
- multi class
- active learning
- dynamic time warping
- gaussian processes
- learning problems
- covariance matrix
- dimensionality reduction methods
- principal component analysis
- similarity measure