WCCI Performance Prediction Challenge

The challenge is over, but a new challenge is on-going using the same datasets, check it out!

SVM-CV

Submitted by Olivier Chapelle

RBF Kernel, model selection by minimzing 2-fold CV
Performances estimated with Leave-One-Out.

Dataset Balanced Error Test guess Guess error Test score Area Under Curve
Train Valid Test Train Valid Test
ada 0.1349 0.1799 0.1863 0.1675 0.0188 0.2051 0.9417 0.9112 0.8989
gina 0 0.0569 0.0685 0.0723 0.0038 0.0721 1 0.9794 0.9799
hiva 0.035 0.3763 0.3049 0.2273 0.0776 0.3825 0.9998 0.6149 0.7369
nova 0.002 0.088 0.0749 0.0787 0.0038 0.078 1 0.9653 0.9737
sylva 0.0042 0.0154 0.0136 0.0099 0.0037 0.0172 0.9998 0.9991 0.9987
Overall 0.0352 0.1433 0.1296 0.1111 0.0215 0.151 0.9883 0.894 0.9176

This entry is a complete valid challenge entry.