The prediction on the election of representatives

Abstract

The Senate and House of Representatives (SHR) are the decision-making departments of its national policy and development strategy. It is very significant to predict the election of SHR, so that one can understand the political trending of the nation and judge the bilateral relationship with other nations. In this paper, two types of datasets towards SHR election are constructed, including 4 election results of the Senate and the House of Representatives, and the questionnaire data of the senators and representatives collected by the University of Tokyo. Based on the above datasets, this paper conducts experiments on the prediction of SHR election and the analysis via classic methods, involving decision tree model, naive Bayesian classification model, and the support vector machine model. According to the results, the support vector machine model achieves the best performance on the election dataset with the F1 score of 88.11% on the senate election prediction, which will be further improved to 89.37% when combining with the questionnaires data set.