Weightage based clustering

clustering

#1

say I am doing kmeans clustering of products based on volume, revenue,shipsetcnt,salesordercnt,bookingcost.
now if i need to add different weights to different factors based on business need what are the best approaches to follow?

is it possible to build a model that allows end user to provide weights before running the model.

this is unsupervised learning


#2

@rumsinha,

After normalizing the data by min - max scaling or z-score scaling. You can multiply any weight to the required columns based on business need. Yes it’s possible to provide different weights to the factors.

It is also possible to build a model that allows end user to provide weights before running the model, it can be achieved by using a R shiny dashboard.

Hope this helps.

Regards,
Aayush


#3

Thank You