Using Black Box models in Inspyred

[sambitn...@gmail.com]

Hi team,

Recently I am using a lot of Inspyred for my optimization problems and loving it. I was wondering, can I use a black-box model like xgboost, random forest inside inspyred as my objective function in order to find the optimal input values that maximizes prediction.

I have tried a lot of optimizer but unfortunately most of these either do not allow black-box models as objective or do not allow customized constraints. Let me know what you think.

Thanks,

Sambit

[Aaron Garrett]

As long as you can invoke the black box and then read in the result to determine fitness, then you certainly can do that. I use external simulations all the time as black boxes for optimization. I'm not familiar with your particular case, but I would say it's a virtual certainty that you can do what you want. For instance, if it were a separate program, then you could make your evaluator function start a process that runs the program and then reads standard output (or an output file if the program generates that) to get the output from which to calculate the fitness.

--
Aaron Garrett

--
You received this message because you are subscribed to the Google Groups "Inspyred" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inspyred+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/inspyred/48012269-1f3f-4161-be0c-9688ccd87f33%40googlegroups.com.