Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/bujowskis/put-ml-spotify-case-study

Repository concerning Machine Learning classes case study - Recommendation of harder styles music
https://github.com/bujowskis/put-ml-spotify-case-study

Last synced: 13 days ago
JSON representation

Repository concerning Machine Learning classes case study - Recommendation of harder styles music

Awesome Lists containing this project

README

        

# put-ML-Spotify-case-study

Repository concerning Machine Learning classes case study - Recommendation of harder styles music

# Introduction

Given the assignment concerning classification of a dataset of our choice, I opted for something that was biting me for a very long time.

Personally, I am quite a music junkie with a rather *particular* taste in music. **Harder styles** such as Techno, Hardstyle, Hardcore, Gabber, Uptempo, etc. are one of my favorites and take up a big part of my music consumption.

I'm extensively using Spotify as a music streaming service of my choice. The problem is, the recommendations system that's supposed to provide me with new tracks I will like does a pretty bad job at it - my **Discover weekly** playlist has an accuracy of just `1/30` songs.

I'm unsure whether that's a problem with Spotify's system being bad or me being too picky. That's why in this case study I'd like to explore that problem by applying different classifiers to see if they can do better, thus inducing that the current Spotify's system could be improved.

# Presentation
You can find the presentation in `.html`, `.Rpres` and `.md` format in the [following folder](https://github.com/bujowskis/put-ML-Spotify-case-study/tree/main/presentation). There's no .pdf version, since the presentation contains gifs.

It's most likely the most convenient for you to view the [Markdown format version](https://github.com/bujowskis/put-ML-Spotify-case-study/blob/main/presentation/cs-presentation.md)