Predicting Box Office Receipts Of Movies With Pruned Random Forest

Predicting box office receipts of movies in theatres is a difficult and challenging problem on which many theatre managers cogitated.The theater manager through the estimation of a movie to decide the time of release film program.Good schedule can greatly improve the cinema box office revenue,reduce the operation cost of the studios,so row piece for studios business is very important,for the theater row piece is productivity.In this study,we use pruned random forest to predict the box office of the first week in Chinese theatres one month before movies' theatrical release.Random forest is a combination of several decision tree predictors in machine learning.To classify a new object from an input vector,we put the data into all trees in the forest.Each tree gives a classification result and votes for the predicted class.The forest chooses the class with the most votes.In this paper,we prune random forest through its strength and correlation.And the pruned random forest outperforms conventional random forest.In our model,the prediction problem is converted into a classification problem,where the box office receipt of a movie is discretized into eight categories.In predict system,the box office data was supplied by 68 Chinese theatres and the movie information is downloaded from diffierent movie websites.Experiments on 68 theatres show that the proposed method outperforms other statistical models.In fact,our model can predict the expected revenue range of a movie,it can be used as a powerful decision aid by theatre managers.