DMCA
Convex Sparse Coding, Subspace Learning, and Semi-Supervised Extensions
Cached
Download Links
| Citations: | 10 - 3 self |
BibTeX
@MISC{Zhang_convexsparse,
author = {Xinhua Zhang and Yaoliang Yu and Martha White and Ruitong Huang and Dale Schuurmans},
title = {Convex Sparse Coding, Subspace Learning, and Semi-Supervised Extensions},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Automated feature discovery is a fundamental problem in machine learning. Although classical feature discovery methods do not guarantee optimal solutions in general, it has been recently noted that certain subspace learning and sparse coding problems can be solved efficiently, provided the number of features is not restricted a priori. We provide an extended characterization of this optimality result and describe the nature of the solutions under an expanded set of practical contexts. In particular, we apply the framework to a semisupervised learning problem, and demonstrate that feature discovery can co-occur with input reconstruction and supervised training while still admitting globally optimal solutions. A comparison to existing semi-supervised feature discovery methods shows improved generalization and efficiency.
Keyphrases
convex sparse coding semi-supervised extension subspace learning optimal solution semi-supervised feature discovery method extended characterization improved generalization feature discovery classical feature discovery method automated feature discovery expanded set certain subspace learning sparse coding problem learning problem input reconstruction practical context fundamental problem machine learning optimality result
