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.

vn5

Submitted by Vladimir Nikulin

vn5

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.1637 0.1674 0.1751 0.8607 0.8425 0.8331 agnostic
gina 0 0 0.0535 1 1 0.9464 agnostic
hiva 0.1633 0.1695 0.3012 0.9147 0.8846 0.7348 agnostic
nova 0.0004 0 0.0471 1 1 0.9456 agnostic
sylva 0.0064 0.0041 0.0115 0.9943 0.9996 0.9858 agnostic
Overall 0.0668 0.0682 0.1177 0.9539 0.9453 0.8891 agnostic

This entry is a complete agnostic learning entry.