Model selection

lmsasu

Hello everyone,

The x-validation operator is used in RM to assess the model's performance. Is there any approach to use croos validation for model selection? E.g. choosing the number of neurons in a MLP, or other customizable parameter? Here I refer to some of the approaches described in http://arxiv.org/PS_cache/arxiv/pdf/0907/0907.4728v1.pdf.

Thanks,
Lucian

land

Hi Lucian,

yes this is indeed possible. See the Sample processes that come with RapidMiner to get to know how. I would suggest taking a look at the 06_Meta directory.

Greetings,
Sebastian