摘要
研究了一种用于图像与模式识别的小波神经网络模型 ,给出了相应的算法和计算公式 ,并进行了仿真模拟 .该模型克服了传统 BP网络隐层单元数目难以确定、收敛速率较慢以及易于收敛到局部极小点等缺点 .仿真结果表明网络性能和收敛速率均明显优于传统 BP网络 。
A wavelet neural network used in image and pattern recognition was studied. The algorithm and formulas were presented, and the simulated experiments were carried out. The result shows that this model can overcome the shortcomings of BP networks, such as the uncertain unit number of the hidden layer, the slowly learning rate, and easy to converge to the local minima. All these remarkable features enable the new model to be of good prospects in application.
出处
《中国矿业大学学报》
EI
CAS
CSCD
北大核心
2002年第5期382-384,389,共4页
Journal of China University of Mining & Technology
