US-2016-Presidential-Polls-Analysis-

Worked on FiveThirtyEight 2016 presidential polling dataset, to build a model which would predict the winner accurately. Using Cross Validation ,trained three optimal classification models namely artificial neural network, support vector machines and naive Bayes. Compared the three models using Machine learning Benchmark package(caret) and selected SVM model as it was successful in predicting the winner of 2016 presidential elections with an accuracy of 85%.