angel b.
unread,May 16, 2023, 10:47:59 AM5/16/23Sign in to reply to author
Sign in to forward
You do not have permission to delete messages in this group
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to HeuristicLab
In Genetic programming - Symbolic classification we perform an experiment with 1000 runs.
We then analysed the results with: Executions / Analyse executions / Variable impacts.
In Variable impacts there are several columns referring to each input variable (the relative frequency of the variable in each run, the mean, the standard deviation and the p-value).
I am using the p-value to perform a variable selection and I need to document some aspects of its use in HeuristicLab. I have searched the documentation, but have been unable to find an answer. I would be grateful if you could clarify several aspects regarding this p-value
What is the null hypothesis - H0? I guess something about the frequency of the variable in the experiment being due to chance?
What is the distribution function of this H0? Binomial, Poisson, Normal...?
What are the parameters - average and variance - of the distribution function of H0?
What is the exact calculation algorithm?
Thank you in advance for your attention
Best, Ángel