基于小波采样理论的新型准则函数.pdfVIP

第

53

卷第

1

期电子科技大学学报

Vol.

53No.

1

2024年1月Journal

of

University

of

Electronic

Science

and

Technology

of

China

Jan.

2024

基于小波采样理论的新型准则函数

12*1

毛伟伟，张治国，金晓宇

（1.

海军士官学校

5系，蚌埠

233012；2.

电子科技大学

自动化工程学院，成都

610054）

摘要为解决在噪声环境下建模的过拟合问题，基于小波采样理论，提出一种适用于小波神经网络的新型准则函数，并

设计了相应的训练算法。这种算法能够利用样本分布和误差训练输入和输出层权值，因此可以大大提高小波神经网络的学习

效率。理论和试验表明，新型准则函数有力地保证了小波神经网络的泛化能力，其相应的算法具有全局收敛性，并对噪声变

化具有良好的鲁棒性。

关键词广义采样；神经网络；过拟合；小波采样

中图分类号TN911.97

文献标志码A

DOI10.12178/1001-0548.2022128

A

Novel

Cost

Function

Based

on

Wavelet

Sampling

Theory

12*1

MAO

Weiwei,

ZHANG

Zhiguo,

and

JIN

Xiaoyu

(1.

Fifth

Department,

Naval

Petty

Officer

Academy,

Bengbu

233012,

China;

2.

School

of

Automation

Engineering,

University

of

Electronic

Science

and

Technology

of

China,

Chengdu

610054,

China)

AbstractIn

order

to

solve

overfitting

of

modeling

in

noisy

circumstance,

a

novel

cost

function

with

corresponding

training

algorithm

is

proposed

for

wavelet

networks

based

on

sampling

theory.

Since

such

an

algorithm

can

use

sample

distributions

and

errors

respectively

to

train

input

and

output

weights,

learning

efficiencies

of

wavelet

networks

are

improved

greatly.

The

theories

and

experiments

show

that

this

novel

cost

function

can

ensure

generalizations

of

wavelet

networks.

Simultaneously,

the

new

algorithm

can

conv