概率神经网络回归、判别与聚类

Probabilistic Neural Networks for Regression, Decision and Clustering

本文基于相同结构的概率神经网络,提出了使其具有回归、判别与聚类功能的不同权重学习算法以及通用变元最优选择方法.该网络具有结构和学习算法简单、收敛速度快、易于软件模拟实现等优点,并可方便地应用于预报、控制、决策等领域.

Based on probabilistic neural network(PNN),the learning algorithms for regression decision and clustering are presented in this paper.The general method for optimal variable selection is also provided.The advantages of PNN are that it is easy to be implemented with highly parallel structure,to be trained with simple algorithms and to be simulated with software.This network can be widely used in pattern recognition,prediction,control and modeling.