End-to-end Kmeans Cluster analysis in R with a Customer Segmentation Project

Download here the Cluster Analysis script and the data for this project: data-heroes-2.ck.page/cluster...
To access my secret discount portal:
linktr.ee/diogoalvesderesende
Customer segmentation is one of the key components of any marketing plan. Arguably, a great segmentation addresses the following questions:
1) Who are the target customers?
2) How to serve them?
3) Where do we need to play?
To be able to answer these 3 questions, the segmentation has to be (1) actionable, (2) differentiable, and (3) have an economic perspective in mind.
In this Case Study, I segment customers based on demographic characteristics and personal characteristics, totaling 16 in the end. The technique of choice is Kmeans, one of the most simple clustering techniques. The output is 6 characteristics that are interpretable, actionable, and that takes an economic perspective mind. How? Be sure to watch the video and find out.

[@DataHeroes]

Download here the Cluster Analysis script and the data for this project: data-heroes-2.ck.page/cluster_analysis

[@AayushKumarak2401]

A very good mix of tech and stats without going too much into any of it. Very well done

[@jorgedias7238]

The best Kmeans example I've seen on KZbin. The majority of practical videos uses Iris Dataset (which only as nominal variables). This one has nominal variables, boolean, nominal with nule values, etc. Keep up the good work!

[@AsmaaSabiri]

Thanks a lot. Very well Explained.

[@DataHeroes]

Thanks!!

[@thiendinh4632]

Thank you a lot, very useful example. Can I know how can we choose the range of cutoff ( example is -0.35 and 0.35 )?

[@jorgedias7238]

At 16:45, when normalizing, why did you want to create a data frame instead of a matrix? Only for simplifying visualization purpose? Would the rest of calculation be different if you had done dataset

[@SimoBenziane]

This is amazing, Thank you. very on point explanation. Question: Can it be in a dashboard as a Webapp? to get Live information from a cllient and decide to which cluster they belong to ?

[@SimoBenziane]

@@DataHeroes I found one video doing exactly that: kzbin.info/www/bejne/e3SYZJmqp76Zb6M Your video helped me a lot thank you tons

[@romainvanderelst3226]

May I ask you which version of Rsudio are you using?

[@anacarolinaaraujo5313]

Hi there. Have you managed to run all the functions? mine crashed at the end (plotting)

[@kirstenkeith9427]

Why did you scale the dataset on 1:16?

[@DataHeroes]

So that all variables have similar weights during clustering