Giters

ClairHu / deezer_report

DSG17 | International Machine Learning Competition from Deezer | The goal of this challenge was to predict whether the users of the test dataset listened to the first track of Deezer's own music recommendation algorithm proposed them or not.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

deezer_report

Data Science Game 2017 | Online Qualification Phase: Music recommendation

  • Started: 12:22 am, Friday 14 April 2017 UTC
  • Ended: 11:59 pm, Wednesday 31 May 2017 UTC (47 total days)

Deezer is a music streaming app, also available on the web. It proposes more than 43 million tracks and is available in more than 180 countries, through a free limited service and a premium offer.

This was international student competition where more than 220 universities and around 800 students participated in online qualification round which lasted for 47 days. This competition was hosted on Kaggle. We participated in this competition as a team of 4 students from our university - Berlin School of Economics & Law. https://www.kaggle.com/teamten

Main files:

Statistics about this competition:

alt text

The goal of this challenge was to predict whether the users of the test dataset listened to the first track of Deezer's own music recommendation algorithm proposed them or not. The evaluation metric for this competition was the ROC AUC.

Results:

  • Rank in Germany by Private Leaderboard Score alt text

Top 20 teams/ Finalists

alt text

About

DSG17 | International Machine Learning Competition from Deezer | The goal of this challenge was to predict whether the users of the test dataset listened to the first track of Deezer's own music recommendation algorithm proposed them or not.