Movie Box Office Prediction System Based on Multi-Architecture Neural Network and Fish School Algorithm

Abstract

In recent decades, the film industry has developed rapidly, both in terms of the number of films and the production cost, showing an increasing trend year by year. The huge scale of the film market and high film production costs have made box office forecasts an urgent concern for the film industry. As early as the 1980s, the industry began to study the problem of movie box office forecasts. Throughout the empirical analysis of the relevant reflection that effect the movie box office, the box office prediction model is constructed. This paper builds a movie box office prediction system based on multi-architecture neural network and fish school algorithm to make intelligent predictions for different types of movies.