ENHANCED TRAINING ALGORITHMS, AND INTEGRATED TRAINING ARCHITECTURE SELECTION FOR MULTILAYER PERCEPTRON NETWORKS

作     者：BELLO, MG 

作者机构：Charles Stark Draper Laboratories Inc. Cambridge MA USA 

出 版 物：《IEEE TRANSACTIONS ON NEURAL NETWORKS》 (IEEE Trans Neural Networks)

年 卷 期：1992年第3卷第6期

页      面：864-875页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 0808[工学-电气工程] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

主　　题：Multilayer perceptrons Integrate Backpropagation algorithms algorithms pattern recognition problem SYNTHETICS Nonlinear least squares 

摘      要：The standard backpropagation based multilayer perceptron training algorithm suffers from a slow asymptotic convergence rate. In the work reported here, sophisticated nonlinear least squares and quasi-Newton optimization techniques are employed to construct enhanced multilayer perceptron training algorithms, which are then compared to the backpropagation algorithm in the context of several example problems. In addition, an integrated approach to training and architecture selection that employs the described enhanced algorithms is presented, and its effectiveness illustrated in the context of synthetic and actual pattern recognition problems.