Implementation issues of sigmoid function and its derivative for VLSI digital neural networks
- 1 January 1992
- journal article
- Published by Institution of Engineering and Technology (IET) in IEE Proceedings E Computers and Digital Techniques
- Vol. 139 (3), 207-214
- https://doi.org/10.1049/ip-e.1992.0033
Abstract
This paper proposes a number of different implementations for the first derivative of the sigmoid function. The implementation of the sigmoid function employs a powers-of-two piecewise linear approximation. The best implementation scheme for the derivative is suggested based on overall speed performance (circuit speed and training time) and hardware requirements.Keywords
This publication has 6 references indexed in Scilit:
- Introduction to the Theory of Neural ComputationPhysics Today, 1991
- Fast Generation of Neuro-ASICsPublished by Springer Nature ,1990
- Explorations in Parallel Distributed Processing - Macintosh versionPublished by MIT Press ,1989
- Why VLSI implementations of associative VLCNs require connection multiplexingPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1988
- Serial-Data ComputationPublished by Springer Nature ,1987
- Digital Transmission SystemsPublished by Springer Nature ,1985