Bollywood Movie Success Prediction using Machine Learning Algorithms

Hindi Film Industry also referred to as Bollywood has now become a multibillion dollar industry and has also surpassed Hollywood in terms of amount of ticket annually sold. With so much money now riding on Bollywood movies it has become imperative to make accurate predictions about success of Bollywood films Today even if a Bollywood movie does not become a hit at box-office the producer of movie will still make profits through sale of satellite rights and music rights but it is Distributors who suffer losses. Hence it has now become imperative for distributors to purchase distribution rights of movies at reasonable prices such that they can obtain profits from it rather than just break even. This problem is a supervised learning problem and will review regression techniques discussed in the literature for predicting the lifetime net India collections of Bollywood films as well as use classification methods on our Bollywood movie dataset for multiclass classification. An evaluation of all the approaches is proposed in which the accuracy score will be reported.