4366 search results - page 4 / 874 » Learning and Smoothed Analysis |
ML
2007
ACM
13 years 7 months ago
2007
ACM
In many learning problems prior knowledge about pattern variations can be formalized and beneficially incorporated into the analysis system. The corresponding notion of invarianc...
COLT
2008
Springer
13 years 9 months ago
2008
Springer
We study the potential benefits to classification prediction that arise from having access to unlabeled samples. We compare learning in the semi-supervised model to the standard, ...
TNN
1998
13 years 7 months ago
1998
— In this paper, we perform a complete asymptotic performance analysis of the stochastic approximation algorithm (denoted subspace network learning algorithm) derived from Oja’...
ENVSOFT
2000
13 years 7 months ago
2000
A technique is proposed for smoothing a broken line fit, with known break points, to observational data. It will be referred to as "broken line smoothing". The smoothness...
ICPR
2010
IEEE
13 years 9 months ago
2010
IEEE
Canonical correlation analysis (CCA) is a powerful tool for analyzing multi-dimensional paired data. However, CCA tends to perform poorly when the number of paired samples is limit...