An Expert Cognitive System Using Ada-boost Algorithm

Recent works on ensemble methods like Adaptive Boosting have been applied successfully in many problems. Ada-Boost algorithm running on a given weak learner several times on slightly altered data and combining the hypotheses in order to achieve higher accuracy than the weak learner. This paper presents an expert system that boosts the performance of an ensemble of classifiers. In, Boosting, a series of classifiers is used to predict the class of data where later members of the series concentrate on the training data that is incorrectly classified by the previous members. To classify an instance, each classifier predicts the class of a new instance and these predictions are then combined using majority voting. Then the class of a new instance with highest weighted vote is considered. We have used the naïve-Bayesian classifier as the weak learner for the expert system. Extensive empirical results are provided using 20 real world dataset and compared with its own weak learner naïve-Bayesian classifier.