Home

Awesome

NTDS project: Transitional playlists for new musical discoveries

This project is done in the frame of the EPFL course Network tour of data science. The main goal is to take some of the most listened tracks found in the Free Music Archive and help someone wishing to discover a new style or a new artist to transition smoothly from a song he knows and likes to an other unknown music. Basically the idea is to let the user choose two points in a graph representing the music tracks and return him a 10(max)-tracks playlist gradually changing from musics similar to the first track to others more like the the second track.

Code and notebooks

Dependencies: