PASCAL Agnostic Learning
vs.
Prior Knowledge
IJCNN07

The challenge is now over. But it remains open for post-challenge submissions!


IMPORTANT: Entries made since February 1st 2007 might be using validation data, now available for training.

mixed ensembles

Submitted by Joerg Wichard

mixed ensembles - mainly trees - balanced data sets - unscaled
no feature selection

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.163 0.2013 0.1786 0.9212 0.8771 0.9067 agnostic
gina 0.0124 0.0667 0.0707 0.9995 0.9878 0.9795 agnostic
hiva 0.1666 0.3015 0.3265 0.9252 0.7523 0.7451 agnostic
nova 0.0982 0.134 0.1326 0.9778 0.9358 0.9423 agnostic
sylva 0.0047 0.0111 0.0072 0.9996 0.9989 0.9987 agnostic
Overall 0.089 0.1429 0.1431 0.9647 0.9104 0.9145 agnostic

This entry is a complete agnostic learning entry.