模糊状态风险分析的广义Logistic回归理论与应用(1)——模糊水平聚类和变量筛选通用算法

The Theory and Applications of Generalized Logistic Regression for Risk Analysis of Fuzzy State(1)—A General Algorithm for Clustering Fuzzy Levels and Selecting Covariates

摘要: 论述了“微弱相关影响因素”概念及其在预防医学中的重要意义,并研究了适宜处理此类资料的统计方法.借助模糊状态概念、交叉积差和统计量和信息量寻优标准,发展了暴露水平聚类优化、状态变量选择通用算法,可明显提高检测、识别微弱相关影响因素的效能与统计分析水准.

Abstract: The concept and statistical methods for weakly correlative factors,influencing the formation and development of chronic diseases,are proposed in this paper.By the aid of fuzzy state analysis, the Cross-Product Difference Sum(CPDS)and the Akaike's Information Criterion(AIC),a general algorithm,clustering fuzzy exposure levels and selecting covariates,is designed for solving problems connected with analytical abilities of detecting and recognozing weakly correlative-influencing factors.