ensongble

Machine learning project to predict the genre of a movie using the movie's soundtrack. Stemmed from an interest in the added storytelling power of music and the scoring of movies.

process_data.py

• loads audio buffers from audio files
• serializes audio buffers by movie
• serializes features
• serializes final dataframe

generate_production_model.py

• generates the production model to deploy using dataframe created by process_data script

train_test_by_song.py

• train and test predictions of songs' genres

train_test_ensemble.py

• train and test using an ensemble approach for the predictions of movies' genres

scraper.py

• gets the top 200 movies' names and meta data from each movie genre from IMDB