Framework for Constructing Hybrid Classifier Using Weight Learning to Combine Heterogeneous Classifiers

In this paper, we present a weight learning method introduced to learn weights on each individual classifier to construct an ensemble. Genetic algorithm is applied to search for an optimal combination of weights for each individual classifier on which classifier ensemble is expected to give best performance. Our proposed ensemble approach can combine heterogeneous classifiers and/or classifier ensembles to enhance the overall classification performance of a given classifier system. We have evaluated our proposed ensemble approach on variety of real life datasets. The proposed approach is compared with existing state-of-the art ensemble techniques such as Adaboost, Bagging and RSM to demonstrate the superiority of proposed work as compared to the competitors.