摘要
随着生活水平的提升,电影作为一种娱乐消费品,已经成为人们生活之中不可或缺的商品,因此对电影票房的预测有着重要的意义。文章将基于多元回归和深度学习设立模型,把2015和2016两年的票房前25部作为研究对象,随机抽取10部作为预测集,其它40部作为训练集,从而分析得出影响电影票房的相关因素以及预测电影票房的最佳模型。
With the improvement of living standards, films, as recreational products, have become an indispensable part of people＇s lives,and hence the film box office forecast is of great significance. In this paper,we will set up a model based on multiple regression and deep learning and use the top 25 movies on the box office standings in 2015 and 2016 as the research object,randomly selecting 10 movies as the forecast set and the other 40 movies as the training set to analyze the factors impacting the film box office and work out the best model to predict movie box office.
出处
《湖北第二师范学院学报》
2017年第8期60-64,共5页
Journal of Hubei University of Education
关键词
预测
票房
多元回归
深度学习
forecast
box office
multiple regression
deep learning
