Mixture weight influence on kernel entropy component analysis and semi-supervised learning using the Lasso

The aim of this paper is two-fold. First, we show that the newly developed spectral method known as kernel entropy component analysis (kernel ECA) captures cluster structure, which is very important in semi-supervised learning, and we provide an analysis showing how mixture weights influence kernel...

Full description

Saved in:
Bibliographic Details
Main Authors: Myhre, J. N., Jenssen, R.
Format: Conference Proceeding
Language:English
Subjects:
Cluster assumption
Clustering algorithms
Convolution
Data spectroscopy
Eigenvalues and eigenfunctions
Entropy
Heart
Kernel
Kernel entropy component analysis
Lasso
Mixture densities
Semi-supervised learning
Spectral dimensionality reduction
Vectors
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The aim of this paper is two-fold. First, we show that the newly developed spectral method known as kernel entropy component analysis (kernel ECA) captures cluster structure, which is very important in semi-supervised learning, and we provide an analysis showing how mixture weights influence kernel ECA in a mixture of cluster components setting. Second, we develop a semi-supervised kernel ECA classifier based on the Lasso framework, and report promising results compared to the state-of-the art.
ISSN:1551-2541
2378-928X
DOI:10.1109/MLSP.2012.6349814