/SpotifyGenre-classification

Multi label classification of Spotify music

Primary LanguagePython

SpotifyGenre-classification

Multi label classification of Spotify music

Description

Multilabel classification of music genre using machine learning and deep learning algorithms. Three different models are trained on processed data. Deep neural network seems to outperform randomforest classifier and logistic regression with chain mechanism. The overall test accuracy of DNN is 48% whereas test accuracy is 93%. Precision and recall is more than 65%.

Folder structure

Data

Spotify data providing musical features for genre classification. Collected from Kaggle

src

Contains scripts for running models and getting results

Presentation

Power point deck for describing the project

Contributor

Sai Munikoti