摘要
在企业风险等级识别过程中存在风险效应人为等分需要量化处理、多维特征指标识别困难以及不同企业间指标高度相似性等问题.为了更好地解决这些识别问题,将模糊模式识别与神经网络相结合提出模糊模式识别神经网络动态识别方法.该方法既克服了模糊模式识别只能处理静态数据且不能实现特征提取的问题,又克服了神经网络对于高维度、高相似性特征指标识别困难的问题.
When enterprise risks are recognized, the following problems may usually be found. The manmade gradation of its effects needs quantizing; the multi - dimensional characteristic indicators are hard to recognize ; and the heights of indicators among enterprises are similar. In order to solve these problems, this arti- cle puts forth a dynamic recognition method combining fuzzy pattern and neutral network. This combination conquers the flaw of fuzzy system that it can only solve static data rather than extract characteristics, and this method also gets over the difficulties of neutral network in recognizing high - dimensional and highly - similar characteristic indicators.
出处
《辽宁大学学报（自然科学版）》
CAS
2009年第3期255-262,共8页
Journal of Liaoning University：Natural Sciences Edition
基金
集美大学预研基金资助项目(C609001)
关键词
模糊系统
模式识别
神经网络
fuzzy system
pattern recognition
neutral network.
