Research on rough set theory and decision tree method applied to soil evaluation

In this paper, rough set and decision tree combination were used to evaluate the productivity grade of soil in somewhere of Jilin province. The research data had a total of 161 records and 16 attributes. The paper used rough set to reduce the soil attributes, removed 5 redundant attributes and obtained the attributes reduction set, then decision tree method was used to construct the decision tree model, after that classifying rules were withdrawn. The experiment indicates that the data mining methods that unify the rough set theory and the decision tree can remove redundant attributes and retain the internal features of the original data. Compared with the single-use decision tree method, the decision tree scale is smaller and the rule set is more streamlined. The mining efficiency is improved.