WCCI Performance Prediction Challenge

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

SVM-RM

Submitted by Olivier Chapelle

RBF Kernel, model selection by minimizing radius/margin.
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 0.167 0.1783 0.1888 0.1723 0.0165 0.2053 0.9104 0.9068 0.8962
gina 0.0016 0.0506 0.0751 0.0767 0.0016 0.0761 1 0.9783 0.9774
hiva 0.0763 0.3054 0.2902 0.2439 0.0463 0.3364 0.9802 0.657 0.7587
nova 0.0236 0.17 0.1332 0.1333 0.0002 0.1332 0.9925 0.94 0.9359
sylva 0.0053 0.0154 0.0134 0.0095 0.004 0.0174 0.9997 0.9991 0.9987
Overall 0.0548 0.1439 0.1402 0.1272 0.0137 0.1537 0.9766 0.8963 0.9134

This entry is a complete valid challenge entry.