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.

vrs1

Submitted by Vladimir Nikulin

version1

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.1738 0.2175 0.1915 0.8372 0.762 0.805 prior
gina 0 0 0.0268 1 1 0.974 prior
hiva 0.1401 0.1855 0.2833 0.8734 0.7786 0.7322 agnostic
nova 0 0 0.0817 1 1 0.9076 agnostic
sylva 0.0121 0.0098 0.0154 0.9818 0.9864 0.9768 prior
Overall 0.0652 0.0826 0.1197 0.9385 0.9054 0.8791 prior

This entry is a complete prior knowledge entry.