Spotify Most Popular Songs


Business Agenda:

In the competitive space of music, everyone is trying to figure out what the next hit is going to be. Getting ahead of the curve allows executives to make informed business decisions based on what they believe will be profitable. Utilizing aspects of songs such as their beats per minute, the artist, duration and other factors, I have created a machine leaning solution for predicting how popular a particular song based on the aspects of the songs.

Project relating to the data sets containing songs from Spotify

  • Utilized data sets available on Kaggle. One for Visualizations and Larger one for training the model.
  • Checked and ensured the validity of the Data
  • Performed Exploratory Data Analysis on the Data
  • Created a Number of Models
  • Packaged Light GBM Model and Pushed into Production

Data Cleaning and Visualization:
After gathering the data I ran the Exploratory Data Analysis on it and produced the following visualizations:

  • Category: Data Visualization and Regression
  • Date: July 2020
  • Github: View on Github