SVM based Ensemble Learning for Spatial and Temporal Air Pollution Analysis

The paper aims to propose spatial and temporal air pollution data analysis by using Support Vector Machine (SVM) ensemble method along with single SVM, Bagging and AdaboostM1 algorithms to identify three objectives of our research. First, to ascertain a computation method that could handle large amount of spatial and temporal air pollution data. Secondly, to discover a high percentage of classification accuracy of proposed method by deploying ensemble algorithms, i.e. single SVM, Bagging and Boosting algorithms. Thirdly, to find best speed of classification of proposed method in terms of computation time. Finally, we compare the ensemble model accuracy of single SVM, Bagging and AdaboostM1 by using Weka 3.6.5 software in order to choose best model to classify air pollution data with high accuracy.