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.

liknon feature selection + state of art

Submitted by Erinija

Feature selection performed with liknon. Several state of art classification rules tested on selected features. The best based on 10 fold crossvalidation of training data was chosen.

Dataset Balanced Error Area Under Curve  
Train Valid Test Train Valid Test
ada 0.1861 0.1932 0.1852 0.8984 0.8687 0.8975 agnostic
gina 0.1397 0.1459 0.1484 0.9309 0.9351 0.9279 agnostic
hiva 0.2374 0.3042 0.3309 0.8628 0.7338 0.7401 agnostic
nova 0.0268 0.092 0.0725 0.9951 0.9813 0.9814 agnostic
sylva 0.0649 0.0885 0.0699 0.9968 0.9977 0.9967 agnostic
Overall 0.131 0.1648 0.1614 0.9368 0.9033 0.9087 agnostic

This entry is a complete agnostic learning entry.