MEPX 2016.04.03

3 views

Skip to first unread message

Mihai Oltean

unread,

Apr 3, 2016, 4:13:26 AM4/3/16

to Multi Expression Programming

Dear friends,

A new version of Multi Expression Programming X is available for download from www.mepx.org

In this version I have added the Mean Squared Error for the computation of fitness in the case of symbolic regression problems. Mean Squared Error is the sum of squared errors divided by the number of test cases.

Previously the computation of fitness was done only with Mean Absolute Error which is the sum of absolute errors divided by the number of cases.

For classification problems, the fitness is now multiplied by 100 in order to display a true percentage of the incorrectly classified data.

regards,

mihai

Reply all

Reply to author

Forward

0 new messages