🎉Community Raffle - Win $25

An exclusive raffle opportunity for active members like you! Complete your profile, answer questions and get your first accepted badge to enter the raffle.

Join and Win

What model comes out of X-Validation?

Fred12

hi,

I'd like to know what is the model that is going out of X-Validation? is it the model with the best performance, or just the last model? because in 10-fold X-Validation u have 10 models at the end...

furthermore, the same I would like to know from which model the contingency-table is at the output at the end.. is it from the last model, the best model ? because I guess it cannot be the average of all contingency tables...

Find more posts tagged with

AI Studio

Accepted answers

IngoRM

Tom is right, it is the model built on the entire data set. Here is probably a useful discussion which will answer a lot of the question you most likely will have very soon (well, predicting the future is part of our business :smileywink: )

http://community.rapidminer.com/t5/RapidMiner-Studio/What-about-n-models-generated-in-cross-validation-Should-we-not/m-p/31649

Hope this helps,

Ingo

Fred12

ok but I still don't understand how it can be the model that is trained on all the dataset, if we are only training on parts of the dataset...

how is the final model created? first , X-Validation is done as usual, then the final model created from the complete dataset and then outputted? that is what still came not clear to me ....

ok seems its created a n+1th time on all dataset:

if you retrieve the model from the outgoing port of the XValidation, then a model is trained on the complete data set. You will notice that in the status bar: After learning / applying the model n times, it will be learned a n+1 time.
This behavior isn't parameter dependent anymore. It will be produced if the outgoing port is connected (and hence the model will be used later on)

All comments

Thomas_Ott

The model that comes out is a model that is trained on your entire data set.

IngoRM

http://community.rapidminer.com/t5/RapidMiner-Studio/What-about-n-models-generated-in-cross-validation-Should-we-not/m-p/31649

Hope this helps,

Ingo

Fred12

ok but I still don't understand how it can be the model that is trained on all the dataset, if we are only training on parts of the dataset...

how is the final model created? first , X-Validation is done as usual, then the final model created from the complete dataset and then outputted? that is what still came not clear to me ....

ok seems its created a n+1th time on all dataset:

if you retrieve the model from the outgoing port of the XValidation, then a model is trained on the complete data set. You will notice that in the status bar: After learning / applying the model n times, it will be learned a n+1 time.
This behavior isn't parameter dependent anymore. It will be produced if the outgoing port is connected (and hence the model will be used later on)