Current developments such as new effects and 3D shootings increase the competition in the movie industry. Pre-production analyzes are becoming more important for the expensive and risky investments in the movie industry. At this point, the prediction of the box office revenue has become an important research issue. In this context, this study aims to present an approach using machine learning algorithms for box-office revenue prediction. Artificial neural networks and support vector machines algorithms as traditional artificial intelligence methods and random trees, random forests and C4.5 algorithms as decision tree algorithms are used. Later, a hybrid model is proposed using these algorithms and the bagging algorithm from the ensemble algorithm. Prediction models are evaluated with the percentage of correct classification, kappa statistics and ROC area. Numerical results show that Random forest-bagging and artificial neural networks-bagging hybrid methods have the best performance among all models.
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|