ON THE PROS AND CONS OF DIFFERENT AVAILABLE ALGORITHMS

[Giulia Ulpiani]

Dear all, 

I was wondering if you ever compared the pros and cons of different algorithms included in HeuristicLab to serve as a guideline for users.

Thank you

[Gabriel Kronberger]

Hi,

for derivative-free optimization of continuous problems I like to use CMAES. For binary problems the parameterless population-pyramid (P3) algorithm is a good choice. For regression problems the choice is determined by the type of model you want to have (e.g. sparse linear model -> use elastic-net). If you don't care about the regression model structure random forest or gradient boosted trees work well.

Best, Gabriel

--

You received this message because you are subscribed to the Google Groups "HeuristicLab" group.

To unsubscribe from this group and stop receiving emails from it, send an email to heuristiclab...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/heuristiclab/14cf31bd-7891-4f74-9418-18a764196a6co%40googlegroups.com.