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.

svm

Submitted by mathias

svm

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.172 0.1868 0.1821 0.8275 0.8288 0.8178 agnostic
gina 0 0.0319 0.0581 1 0.9723 0.9425 agnostic
hiva 0 0.2676 0.3226 1 0.7193 0.6814 agnostic
nova 0.0054 0.036 0.0538 0.9949 0.9502 0.9473 agnostic
sylva 0.0073 0.0115 0.0225 0.9931 0.9884 0.9775 agnostic
Overall 0.0369 0.1067 0.1278 0.9631 0.8918 0.8733 agnostic

This entry is a complete agnostic learning entry.