I created a model from data on over 7300 movies from TMdB, and used it to predict box office revenues for a Kaggle competition. An interative version is deployed on Heroku, and the notebook is on my GitHub.