This python project explores machine learning and data visualization to analyze a Yandex Music dataset of song attributes to implement a Decision Tree.
Using song attributes data, I implemented an algorithm to predict whether or not the user will like or dislike a song.
• I explored a dataset on kaggle which provided a list of a user's songs that were labled with a target value of either 0 or 1 depending on whether or not the user liked or disliked a song
• A features list was created to store different attributes that YMusic extracts to make up the skeleton of a song (e.g duration, loudness, tempo, etc)
• A series of histograms were created for each attribute, which were split on the target value (1 or 0) of what the user liked and disliked about a song
• I analyzed the distribution of value for particular features (e.g the user had a preference for songs with a higher danceability lower loudness)
• I then built a machine learning model to implement a decision-tree classifier and visualized it