Research On The Correlation Analysis Of Forest Subcompartment Factor Based On Chi Square Test

Forest resources subcompartment data had the characteristics of volume was so great,factor relationship was so complicated,data types was so various, speed changing was so high,density value was so low, which catained more than 100 factors, there were forestland type,forest category, origin, forest class, land ownership, soil type, traffic location, gonernance level and so on, these factors had classification variables and continuous variables. How to extracted strong correlation factor combination quickly and mined the correlation between factors, which is the efficient way to concentrated the forest resources subcompartment data, observed emphatically and optimized the statistical report forms, and extracted valuable information and the rule of deep and knowledge from the forest resources subcompartment data.But there are no suitable methods so far.This paper was based on the theory of data mining technology and chi-square test method,analyzed the factors correlation of forest resources subcompartment data, in order to find strong correlation factor combination;Besides, estimated the correlation index of classification variables and analyzed the strong correlation factor combination, digged up the strong correlation between the fator-members which belongs to the factor combination.Here is the main research content in this article:(1) Research on the connotation of factors correlation between forest resources subcompartment data. Analyzed aspects of the forest resources subcompartment data and its relationshipbetween factors,correlation degree evaluation standard,the application characteristics of correlation analysis, etc. Expounded the factors connotation of the forest resources subcompartment data, laied the foundation for subcompartment factor correlation analysis based on chi-square at the same time.(2) Builded model for factor correlation based on chi-square. Based on chi-square algorithm of correlation analysis, through modeling description link of the data standardization,the modeling process, correlation analysis and calculate, completed the factor relationship modeling of the forest resources subcompartment data.(3) Experimented and analyzed of factor correlation based on the chi-square. Based on previous factor correlation content research and factor correlation model, used the forest resources subcompartment data of Liao Ning province, extracted the fator-members and factor combination which had strong correlation, and put into use it in the management of forest resources statistics optimization, afforestation design and logging design to find the potential valuable knowledge.