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.

cv-trees

Submitted by Joerg Wichard

an ensemble of cross-validated decision trees,
the first trial.

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.2458 0.2586 0.2482 0.8883 0.8678 0.8914 agnostic
gina 0.0775 0.0662 0.1001 0.9819 0.9903 0.9618 agnostic
hiva 0.1491 0.133 0.295 0.9414 0.9344 0.7702 agnostic
nova 0.1433 0.128 0.147 0.9237 0.9152 0.9135 agnostic
sylva 0.0324 0.0444 0.0414 0.9995 0.9995 0.9976 agnostic
Overall 0.1296 0.1261 0.1663 0.947 0.9414 0.9069 agnostic

This entry is a complete agnostic learning entry.