/spotify-tracks

Implementação de modelo de Machine Learning para predição de popularidade no lançamento de uma nova música no Spotify.

Primary LanguageJupyter Notebook

spotify-tracks

On this project the ideia is to implement a Machine Learning model that will predict the populary of a track on Spotify.

The project is separated in two main modules:

  • a Node.js service to fetch data tracks from Spotify API and save it on CSV files.
  • a data preprocessing and model evaluation/creation script made on Python using Google Research Colab.

The team members are guivahl and gabrielgomes0.

Node.js Service

When searching for tracks, you may use different filters. We have decided to fetch data from the year of 2021 and also with specific genres (top 10 genres of Brazil).

Installation

Run Locally

  • Install packages with the command yarn
  • Create a .env file and add your Spotify Credentials (you may rename .env.example)
  • Run yarn start

Documentation

Python Colab