An Efficient Algorithm For Predicting The Response Of A Laterally Inhibited Neural Network

B. V.K. Vijaya Kumar; Michael Lemmon

doi:10.1117/12.962372

6 December 1989 An Efficient Algorithm For Predicting The Response Of A Laterally Inhibited Neural Network

B. V.K. Vijaya Kumar, Michael Lemmon

Proceedings Volume 1154, Real-Time Signal Processing XII; (1989) https://doi.org/10.1117/12.962372
Event: 33rd Annual Technical Symposium, 1989, San Diego, United States

Abstract

Neural networks are usually described mathematically using sets of coupled non-linear differential equations. Thus, we can determine the response of the network to any input stimulus by numerically integrating the corresponding differential equations (which can be computationally intensive). In this paper, a new algorithm is presented for determining the input-output behavior of laterally inhibited neural networks. This algorithm is based on an analytical understanding and thus does not involve any numerical integrations. Initial results indicate that this new algorithm provides a speed up of up to three orders of magnitude. We also propose a VLSI Systolic Array implementation of this algorithm.

Citation Download Citation

B. V.K. Vijaya Kumar and Michael Lemmon "An Efficient Algorithm For Predicting The Response Of A Laterally Inhibited Neural Network", Proc. SPIE 1154, Real-Time Signal Processing XII, (6 December 1989); https://doi.org/10.1117/12.962372

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available