Song recommender

Inspiration

Get familiar with machine learning

What it doess

The algorithm learn the user taste based on their playlists and recommends songs

How we built it

We used Spotify's API to build the datasets as well as a neural network of 1 hidden layer.

Challenges we ran into Learning

Learning how to use API's and format the data.

Accomplishments that we're proud of

The algorithm usually had a test accuracy of >90%

What we learned

How machine learning works and some of its applications

What's next for Song recommender

Usage of tensorflow for a neater source code and use of more sophisticated networks for better predictions.

Built With

• python