Research of Box-Office Prediction based on Rough Set and Support Vector Machine

In this paper, a novel prediction method for box-office is proposed based on the rough set data processing function and support vector machine (SVM) classification mechanism. The front-end processor, optimizes the input variables by attribute reduction, in order to improve the performance of classifier. Then, in view of the lack of guidance of scientific theory problem of domestic movie box office prediction, the classifier for the box-office prediction, the influence factors of the box office revenue as the input variables, the boxoffice income categories as output variables, data preprocessing and training test. Results show that the classifier can effectively solve the box office prediction problem, the results of the multilayer perceptron is better than that of Ramesh S. and Dursun D. using the prediction method, and the prediction error is less than 10%, to meet the requirements of the film market, show the powerful classification ability.