Oscillatory Neural Networks for Pattern Recognition
Author | : Vonda H. Miller |
Publisher | : |
Total Pages | : 198 |
Release | : 2006 |
ISBN-10 | : 1109889283 |
ISBN-13 | : 9781109889284 |
Rating | : 4/5 (284 Downloads) |
Download or read book Oscillatory Neural Networks for Pattern Recognition written by Vonda H. Miller and published by . This book was released on 2006 with total page 198 pages. Available in PDF, EPUB and Kindle. Book excerpt: Computer algorithms that match human performance in recognizing written text or spoken conversation remain elusive. The reasons why the human brain far exceeds any existing recognition scheme to date in the ability to generalize and to extract invariant characteristics relevant to category matching are not clear. However, it has been postulated that the dynamic distribution of brain activity (spatiotemporal activation patterns) is the mechanism by which stimuli are encoded and matched to categories. This research focuses on modeling and analyzing dynamic patterns for category discrimination. It is assumed in this research that EEG patterns are related to perception. It is further assumed that the distribution of brain activation patterns are self-emergent and this self-organization is sufficient for categorization. These assumptions have loose biological underpinnings; nonetheless, there is evidence to support these premises. An oscillatory neural network has been designed based upon a canonical model. The results of experiments for setting the parameters of this neural network are presented along with the emergent spatiotemporal activation patterns to simple stimuli. Classification of spatiotemporal frequency transitions and their relation to apriori assessed categories is shown. Classification is accomplished using a trajectory based distance metric. Since the distance metric is differentiable, a supervised learning algorithm based on gradient descent is demonstrated. The results indicate that this spatiotemporal representation of stimuli and the associated distance metric is useful for simple pattern recognition tasks.