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.

vrs2

Submitted by Vladimir Nikulin

version2

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.1784 0.1948 0.188 0.8503 0.8114 0.8287 prior
gina 0 0 0.0266 1 1 0.975 prior
hiva 0.1553 0.1465 0.2931 0.9085 0.8886 0.7491 agnostic
nova 0 0 0.0802 1 1 0.9187 agnostic
sylva 0.013 0.0156 0.0157 0.9794 0.9742 0.9762 agnostic
Overall 0.0694 0.0714 0.1207 0.9476 0.9348 0.8895 prior

This entry is a complete prior knowledge entry.