The quantization effects of different probability distribution on multilayer feedforward neural networks

Author

Jiang, Minghu ; Gielen, Georges ; Deng, Beixing ; Tang, Xiaofang ; Ruan, Qiuqi ; Yuan, Baozong

Author_Institution

Katholieke Univ., Leuven, Heverlee, Belgium

Volume

2

fYear

2002

fDate

26-30 Aug. 2002

Firstpage

1175

Abstract

A statistical model of quantization was used to analyze the effects of quantization in digital implementation, and the performance degradation caused by number of quantized bits in multilayer feedforward neural networks (MLFNN) of different probability distribution. The performance of the training was compared with and without clipping weights for MLFNN. We established and analyzed the relationships between inputs and outputs among bit resolution, network-layer number, and performance degradation of MLFNN which are based on statistical models on-chip and off-chip training.

Keywords

feedforward neural nets; learning (artificial intelligence); multilayer perceptrons; probability; statistical analysis; MLFNN; bit resolution; clipping weights; digital implementation; multilayer feedforward neural networks; network-layer number; off-chip training; on-chip training; performance degradation; probability distribution; quantization effects; statistical model; Analysis of variance; Degradation; Feedforward neural networks; Gaussian distribution; Multi-layer neural network; Neural networks; Neurons; Performance analysis; Probability distribution; Quantization;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing, 2002 6th International Conference on

Print_ISBN

0-7803-7488-6

Type

conf

DOI

10.1109/ICOSP.2002.1179999

Filename

1179999