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.

Test mixed

Submitted by reference

This example tests that mixed entries are classified as "Prior Knowledge"

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.2273 0.2396 0.2181 0.8687 0.8469 0.8729 prior
gina 0 0.041 0.052 1 0.9954 0.9886 prior
hiva 0.2782 0.3243 0.3197 0.8103 0.7425 0.7349 agnostic
nova 0.1781 0.19 0.2005 0.9657 0.9685 0.9574 agnostic
sylva 0 0.0206 0.0129 1 0.9989 0.999 prior
Overall 0.1367 0.1631 0.1606 0.929 0.9104 0.9106 prior