9th place solution for the WSDM 2019 Spotify sequential skip prediction challenge
Given information about the first half of a Spotify user's listening session, the task is to predict whether each track in the second half of the session will be skipped. I outline my initial approaches taken during the challenge hosted by CrowdAI using gradient boosted trees and long short term memory neural networks, as well as highlight the importance of structuring the user session and track data as correct time series. Given the time limitations of the challenge, I conclude with some suggestions for further research which could be done on this dataset.
This solution placed 9th during the challenge. The user name on the CrowdAI platform is "cddt"