WCCI Performance Prediction Challenge

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

CLOP-models-only-4

Submitted by Juha Reunanen

Same predictions as in CLOP-models-only-3, but with a different guess for the accuracy.
(The BER estimates in submissions 1-3 are - I think - systematically pessimistic compared to actual performance.)

Dataset Balanced Error Test guess Guess error Test score Area Under Curve
Train Valid Test Train Valid Test
ada 0.1837 0.1881 0.2037 0.187 0.0167 0.2204 0.8954 0.8943 0.8825
gina 0 0 0.098 0.0424 0.0557 0.1537 1 1 0.9631
hiva 0.0148 0.0357 0.3678 0.2654 0.1024 0.4702 1 1 0.7118
nova 0.005 0 0.0892 0.0506 0.0385 0.1277 1 1 0.9874
sylva 0.0084 0.0016 0.0341 0.0072 0.0269 0.061 0.9999 0.9999 0.9971
Overall 0.0424 0.0451 0.1585 0.1105 0.048 0.2066 0.979 0.9788 0.9084

This entry is a complete valid challenge entry.