Key influencers - multicollinearity

[stanxl]

Hi everyone,

I am investigating how to use the Key Influencers visual in the services proposed by our company. In that regard, I am especially concerned about how multicollinearity is handled in the calculation leading to the Key Influencers' results.

If anyone also has more information about the algorithm, what kind of models are used depending on the type of data and dependent variable, etc. it would be more than welcome 💡

I already found this article that gives some insight: link

Please note that I am not a statistician or a mathematician, so even though I may be able to handle some of the stats jargon, I am nowhere near to be a specialist.

Thanks!

Stan