Computer Science – Information Theory
Scientific paper
2007-08-29
Journal of Physics Conference Series 95, 012001, 2008; http://www.iop.org/EJ/toc/1742-6596/95/1
Computer Science
Information Theory
Conference paper for the International Workshop on Statistical Mechanical Informatics 2007, September 16-19, 2007, Kyoto, Japa
Scientific paper
10.1088/1742-6596/95/1/012001
A framework to analyze inference performance in densely connected single-layer feed-forward networks is developed for situations where a given data set is composed of correlated patterns. The framework is based on the assumption that the left and right singular value bases of the given pattern matrix are generated independently and uniformly from Haar measures. This assumption makes it possible to characterize the objective system by a single function of two variables which is determined by the eigenvalue spectrum of the cross-correlation matrix of the pattern matrix. Links to existing methods for analysis of perceptron learning and Gaussian linear vector channels and an application to a simple but nontrivial problem are also shown.
No associations
LandOfFree
Inference from correlated patterns: a unified theory for perceptron learning and linear vector channels does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Inference from correlated patterns: a unified theory for perceptron learning and linear vector channels, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Inference from correlated patterns: a unified theory for perceptron learning and linear vector channels will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-240138