基於粗糙集和遺傳算法的多元二次非線性回歸分析模型及應用

摘　要:針對煤與瓦斯突出模式識別是一個非線性、高維數據處理問題，建立了多元二次非線性回歸分析的模式識別模型。利用粗糙集屬性約簡算法對樣本數據進行降維處理，再利用降維數據和遺傳算法估算出模型係數，通過對測試樣本的識別檢驗，表明基於粗糙集和遺傳算法的多元二次非線性回歸分析模型具有很高的判別能力，取得和支持向量機回歸模型一樣的識別效果。

關鍵詞:煤與瓦斯突出 粗糙集 遺傳算法 多元二次非線性回歸分析

分類號:　TD713Based on Rough Sets and Genetic Algorithm of Multivariate Quadratic Nonlinear Regression Analysis Model and ApplicationWANG Jiang-rong , LUO Zi-qin, WEN Hui ( 1. Information Processing and Ccontrol Engineering of Lanzhou Petrochemical Vocational and Technical College,Lanzhou 730060, China; 2. Petroleum Chemical Engineering of Lanzhou Petrochemical Vocational and Technical College ,Lanzhou 730060,China)Abstract:For pattern recognition of coal and gas outburst is a nonlinear, high dimension data processing problems, the establishment of a multiple quadratic non - linear regression analysis model for pattern recognition. Using rough set attribute reduction algorithm was car- ried out on the sample data dimension reduction,reuse data dimension reduction coefficient and the genetic algorithm to estimate the model, through the recognition of test sample inspection, show that based on rough sets and genetic algorithm of multivariate quadratic nonlinear regression analysis model of high discriminant ability, and support vector machine (SVM) regression model as the recognition effect.

Keyword:coal and gas outburst; rough set; genetic algorithm; a multiple quadratic non - linear regression analysis