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.

AUCmax

Submitted by weseeare

BER cross validation

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.2019 0.2329 0.1989 0.9044 0.873 0.9008 agnostic
gina 0.0549 0.1368 0.1692 0.9873 0.914 0.9051 agnostic
hiva 0.1857 0.4197 0.3598 0.9982 0.6002 0.7252 agnostic
nova 0.0056 0.048 0.0519 1 0.9822 0.9869 agnostic
sylva 0.0889 0.0957 0.0878 0.9984 0.9982 0.9976 agnostic
Overall 0.1074 0.1866 0.1735 0.9777 0.8735 0.9031 agnostic

This entry is a complete agnostic learning entry.