With genre prediction we want to classify edm tracks, by its subgenres.
We have created our own dataset using Spotify API, containing tracks from 9 different styles: techno, techhouse, psytrance, trance, hardstyle, ambient, synthwave, trap and dnb.
The first step has been to do an EDA about our data, and proceed with the data cleaning and feature enginereering.
To conclude, we have used machine learning modelling in order to create a classification model, with 95% accuracy.
For the testing, a Streamlit site has been created, which will be online soon.