摘要
针对复杂样本的模式分类问题,提出了有效的因子分析法(FA)、遗传算法(GA)和BP神经网络(BPNN)相结合的FAGABPNN分类方法,基本思想是:首先利用因子分析法对输入样本降维,然后利用遗传算法和BP神经网络相结合的方法对系统进行建模.仿真结果表明,该系统为给复杂样本的分类提供了一条有效的途径.
A complex samples classification system with FAGABPNN was proposed in this paper. To realize the classification system, firstly the dimension of input samples was reduced by Factor Analysis, and then an accurate model was built with Genetic Algorithm and BP Algorithm. Experiment results showed the system was feasible and valid, a valid method for complex samples classification was given.
出处
《重庆工学院学报》
2007年第15期122-125,共4页
Journal of Chongqing Institute of Technology
基金
重庆市教委资助项目(KJ060818
KJ060804)
重庆师范大学校级科学研究项目(06XLB023)
关键词
因子分析
遗传算法
BP神经网络
分类
factor analysis
genetic algorithm
BP neural network
classification
