1Associate Professor, Department of Master of Computer Applications, Siddaganga Institute of Technology, Visvesvaraya Technological University, Tumkur, Karnataka, India
Data mining has great potential for exploring the hidden patterns in different domains. This paper presents the development of an ensemble model, using bagging and multiboost with five base classifiers: RBF, Naïve Bayes, Bayesian, K-nearest classifier and SVM. The development of the model involved two stages: (i) identify the group-sensitive attributes in the large database pertaining to wheat seeds dataset using decision tree and (ii) make a fine-tuned classification using bagging and multi-boost ensemble with five different base classifiers using the sensitive attributes elicited in the first stage. The publicly available wheat seed dataset has been used for the development of the model. Computational experimentations have been conducted using ten-fold cross validations. This investigation conclusively proves the significance of cascading decision tree for feature selection with ensemble methods, namely, bagging and multiboost for the enhanced categorization of wheat dataset.
Feature selection, Ensemble learning, Bagging, Multiboost, Base classifiers, Decision tree
