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.

cross-indexing-prior-1

Submitted by Juha Reunanen

Still using CLOP models only.

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.1333 0.2175 0.1822 0.9476 0.8771 0.9065 agnostic
gina 0 0.0127 0.0267 1 0.9998 0.9964 prior
hiva 0.0801 0.3465 0.2869 0.9701 0.7216 0.7621 agnostic
nova 0.0004 0.062 0.0472 1 0.9936 0.9903 agnostic
sylva 0.0033 0.0033 0.0066 0.9998 0.9995 0.9989 prior
Overall 0.0434 0.1284 0.1099 0.9835 0.9183 0.9308 prior

This entry is a complete prior knowledge entry.