Retro or Not

• 

  Thumbnail

We learned how to use different AWS technologies like Amplify, SageMaker, APIGateway, Jupyter Labs and Lambda for our project.

We used the Spotify API to scrape Spotify playlists of their track information to construct a model of what might be considered ‘retro’. To do this, we wrote a python script that would organize the data points for how a track was retro (danceability, energy, key, loudness, speechiness, acousticness, instrumentalness, liveness, valence, tempo, time_signature). Using SageMaker and XGBoost, we built a machine learning model using these data points to predict whether or not another arbitrary song or playlist is retro or not.

We used Amplify to host our front-end website for inputting spotify playlist links. We used APIGateway for connecting Amplify, Lambda, and Jupyter Labs.

We learned a lot through this hackathon, and we’re excited to continue learning more about machine learning. Sam and Luke primarily focused on the machine learning and backend side of the stack while AT built the front end using HTML, CSS, and Javascript.

Built With

• amazon-web-services
• javascript
• python
• sagemaker
• xgboost