摘要
根据待识别样本的加权广义欧氏权距离,基于模糊距离总变差为最小的目标函数,提出了一种新的计算模糊聚类中心矩阵、模糊模式识别矩阵与目标权重的模糊识别模型.该模型具有较好的稳定性,径流中长期预报实例说明是可行的.
According to the minimum target function of fuzzy distance＇s total variation, a new fuzzy recognition method was presented, which is used to compute fuzzy classification center matrix and fuzzy recognition matrix and objective weight. The model is of the good stability, the applied examples of runoff forecasts in mid-long term shows it＇s feasible.
出处
《数学的实践与认识》
CSCD
北大核心
2009年第2期43-48,共6页
Mathematics in Practice and Theory
基金
国家科技支撑计划项目(2006BAB15B04)
关键词
模糊聚类
模糊模式识别
权重
总变差
径流预报
fuzzy classification
fuzzy recognition
weight
total variation
runoff forecasts
