Approximate Core based RDT model for Classifying Technology Savvy Farmers

摘要

Rough set based Decision Tree (RDT) model is a hybrid of rough set methodology with decision tree induction under the umbrella of data mining applications. RDT model uses a rough set tool called reduct to produce decision trees. A variant of RDT model that emerges while using other rough set tools like core or approximate core instead of reduct is referred as approximate core based RDT model. A case study involving identification of farmers adopting pesticides, insecticides or fungicides in their fields for crop protection is presented in this paper. Results of the experiments using the standard conventional techniques are compared with the hybridized rough set based decision tree approaches. Evaluation parameters accuracy, complexity, number of rules and number of attributes in the classifier were used for comparison. Approximate core based hybridized RDT model is simple and shows accuracy better than or comparable to other classification methods.