Distance-based Bootstrap Sampling in Bagging for Imbalanced Data-Set
G. Rekha, Vuyyuru Krishna Reddy, Amit Kumar Tyagi, Meghna Manoj Nair · 2020 International Conference on Emerging Trends in Information Technology and Engineering (ic-ETITE) · 2020
In the recent decade, many new technologies and problems have attracted attention from research community/scientists. Some problems like imbalanced data-set, security and privacy concerns in various computing environments like internet connected thing (IoTs), cloud computing, distributed computing, etc., have received many innovative ways as efficient answer. But, some problems are still unsolved. Learning from Imbalance dataset with higher accuracy is an essential task/higher priority work in many applications. For handling the class imbalance problems, many extended approaches have been considered for bagging ensembles. In our study we show that application of distance-based approach (DistBagging) for balancing the distribution of each bag in ensemble bagging provides better results for addressing the class imbalance problems. In this work, we propose distance-based approaches for selecting the group of data for each bootstrap method to improve the performance of the classification in terms of accuracy for minority class in the imbalanced class distribution environment. The experimental results show that our distance-based approach outperforms the other ensemble techniques in the previous studies.