Adapting constant multipliers in a neural network implementation

Author

James-Roxby, Philip ; Blodget, Brandon

Author_Institution

Dept. of Electron. & Electr. Eng., Birmingham Univ., UK

fYear

2000

fDate

2000

Firstpage

335

Lastpage

336

Abstract

The use of dynamic reconfiguration appears extremely attractive for implementing adaptive processing algorithms. Often, the adaption involves updating look-up tables based on a parameter which can only be determined at run-time. For reasons of efficiency, these look-up tables are read-only to the rest of the circuitry. This paper compares the use of run-time reconfiguration and read-only look-up tables, with a similar implementation using writable memories. The application under consideration is the multilayer perceptron neural network

Keywords

adaptive systems; multilayer perceptrons; neural net architecture; parallel architectures; random-access storage; read-only storage; reconfigurable architectures; table lookup; RAM; ROM; adaptive processing algorithms; constant multipliers; dynamic reconfiguration; multilayer perceptron; neural network; read-only look-up tables; run-time reconfiguration; writable memories; Artificial neural networks; Biological neural networks; Computer architecture; Intelligent networks; Multilayer perceptrons; Neural networks; Neurons; Read only memory; Runtime; Table lookup;

fLanguage

English

Publisher

ieee

Conference_Titel

Field-Programmable Custom Computing Machines, 2000 IEEE Symposium on

Conference_Location

Napa Valley, CA

Print_ISBN

0-7695-0871-5

Type

conf

DOI

10.1109/FPGA.2000.903442

Filename

903442