WCCI Performance Prediction Challenge

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

GbO+MVf+SVM1

Submitted by Vladimir Nikulin

see for details my paper
"Learning with Mean-Variance Filtering, SVM and Gradient-based Optimization", 8 pages (submitted to WCCI in January 2006).

Dataset Balanced Error Test guess Guess error Test score Area Under Curve
Train Valid Test Train Valid Test
ada 0.163 0.1624 0.1854 0.163 0.0224 0.2077 0.8439 0.8602 0.8146
gina 0 0 0.0567 0.035 0.0217 0.0784 1 1 0.945
hiva 0.2188 0.2293 0.3515 0.232 0.1195 0.4709 0.7782 0.7739 0.6565
nova 0 0.008 0.054 0.045 0.009 0.0629 1 0.9954 0.9303
sylva 0.0119 0.0053 0.0133 0.007 0.0063 0.0197 0.9892 0.9987 0.9862
Overall 0.0787 0.081 0.1322 0.0964 0.0358 0.1679 0.9222 0.9256 0.8665

This entry is a complete valid challenge entry.