Precise Multiplications with the {NEF}

Tech Report, 2015

Jan Gosmann

Abstract

This report discusses how to implement the multiplication of two numbers in the {NEF} with a high accuracy. The main improvement will be achieved by using diagonal encoders instead of randomly distributed encoders. In the selection of the evaluation points a trade-off is found as they cannot give a uniform distribution when projected to the encoders while being limited to the input domain. That leads to an alternative multiplication network architecture improving the accuracy further.

Full text links

PDF

CTN Tech Report

Issn
CTN-TR-20150212-015
Address
Waterloo, ON
Month
2
Institution
Centre for Theoretical Neuroscience

Cite

Plain text

BibTeX