Abstract:
Movie production is one of the most expensive investment fields and can result in enormous financial profit or loss. It is critical for investors and production companies to decide whether to invest in a certain movie given the huge loss that could occur from such investments. Hence, it is very beneficial to construct a model which helps investors in their decision making process. Machine learning has proven its effectiveness in building decision making models and recommender systems in various fields. In this work, we present several machine learning techniques (Support Vectors Machine, K-Nearest Neighbors, C5, Neural Networks and Case-Based Reasoning) along with a genetic algorithm to predict the success of a movie before its production using the IMDB rating as an indicator of the success. Results show that machine learning is useful in this domain and genetic algorithms can be used to build prediction models with relatively good performance.
Keywords:
