Created | Mon, 24 Sep 2012 13:14:57 +0000 |
Published | Wed, 3 Oct 2012 00:53:15 +0000 |
Algorithm | BigML Streaming Tree |
Split criterion | |
Support threshold | |
Depth threshold | 20 |
Node threshold | 512 |
Missing splits | No |
Statistical pruning | Yes |
This model takes the audition results of 5 editions of The Voice (Holland 2010, 2011, Flanders 2011, US 2011, UK 2012) and predicts what the highest level is a contestant can reach. It uses the time it took a first, second, third and fourth coach to turn his/her chair from the moment the vocalist started to sing. If more than 1 coach turned, it also has the time from first to last coach. See this blog post for full details.