Prediction Of TV Play Index Based On Machine Learning Method

In the era of rapid development of the Internet,for the development of film,television production and other industries,massive information data plays a very important role.Through the analysis of these data,we can better target groups to produce films and TV plays,and bring higher profits to enterprises.There is no doubt that big data is playing a more and more important role in the development of film and television industry.At present,there are many researches on the prediction of movie box office based on big data,but the prediction of TV play volume is relatively small.This paper selects the data of local TV series of i QIYI,and uses several methods of machine learning to build models(multiple regression model,Bayesian model,SVM model,decision tree model and random forest model)to predict the TV Drama Broadcast Index.Finally,it is found that the prediction effect of random forest model is the best,and the correct rate is 82.53%.How to choose the type of TV series,how to set the number of TV series,and how many titbits to release;the video broadcast website chooses which kind of TV play to show,and chooses the marketing method;which TV series will advertisers put their advertisements on.It has a certain reference value for the above issues to predict the broadcast index.