Thunderstorm prediction study based on PCA and least square support vector machine

This electronic For the limitations of dependence on previous experience and neural network forecasting model in current thunderstorm prediction. Considering the characteristics of the thunderstorm in Chongqing, the thunderstorm prediction model based on least square support vector machine (LS-SVM) is established. The data are preprocessed by principal component analysis(PCA) firstly. Then, the search space of the penalty parameter and the kernel parameter is defined by analyzing the influence of the two parameters on the performance of LS-SVM classifier and the modeling process and parameters selection are analyzed. Lastly the thunderstorm prediction model based on LS-SVM is constructed and implemented. Comparing with neural network and standard SVM, the results show that the LS-SVM model has better prediction results and faster running speed.