How to apply forest model when there is no predictor variable?

Deepvijay

Hello,

I'm trying to build a model that can split my data into two categories i.e. "Good to Go" & "Not important". As my data do not have any previous responses/column (Response variable) so regular "Supervised" methods won't work for me.

Can anyone suggest how can I utilize unsupervised methods (like, XG Boost, Forest, Boosted Trees) here?

FYI, there is NO predictor variable present in my data.

Thanks,

PhilippK

Hi @Deepvijay ,

check out the following unsupervised clustering methods:

https://community.alteryx.com/t5/Alteryx-Designer-Knowledge-Base/Tool-Mastery-K-Centroids-Diagnostics/ta-p/302311#:~:text=Typically%20the%20first%20step%20of,data%20and%20specified%20clustering%20algorithm.

https://community.alteryx.com/t5/Videos/Cluster-Analysis/td-p/117714

Best regards

Phil