Descriptive Data Mining: K-Means Clustering Using Excel

[@SaraM-c7f]

14:25 can you please explain why it is normalized sales here? what if you have 4 variables in your cluster? thank you.

[@mcanbolat]

All variables should be normalized to make them similar in their magnitudes. Otherwise variables with bigger values would be favored by the model.

[@tyuyhjk3866]

Thank you sir. It was a great tutorial. Kindly upload next tutorial

[@sawekb8102]

I appreciate your work Mustafa. Thank you for shearing. Wish your channel to grow.

[@joserico8939]

excellent video... Question: where can I find the video of the implementation for 3 variables? Thank you so much

[@OmaymanAbdelmajid]

Thanks for this usefull video, how can I download the complete file, the one in the link is not

[@OmaymanAbdelmajid]

How can I get the full file

[@rizalahmadsyarif9191]

Sir you share big skill... You so good person.. 👍

[@hunt_and_gather]

How do you attain the standard deviation for each row? I want to see the formula for column C.

[@mcanbolat]

They are assumed to be given here. If you have a transaction dataset you can use a pivot table to summarize and find the average and standard deviation for each SKU.

[@samrahkhan2803]

Hi. Thanks for this video, it's very helpful. How do you decide constraints to enter into Solver? I'm a little confused.

[@mcanbolat]

Hi, I limit the values to the ranges of the normalized variables. The cluster centers won’t be outside of the min and max values of these variables.

[@aritadasgupta]

very Nice !!

[@richardszalay7608]

Hi! Thanks for your video. :) It gave me useful hints for coding a k-means clustering in VBA. (Just a little thing: the sound level is a little bit low, I had to use a speaker while watching. I don't know the reason, and you might've improved it since then)

[@mcanbolat]

Thank you for letting me know. I realized it later but it was too late. I may have to redo it when I have time.

[@AchooUsman]

Can you share with us data file

[@mcanbolat]

The link is in descriptions.

[@aaliceabuzaid2621]

where does Princiapal component analysis sit here ... i am so mixed up with PCA concept and whether is it fine skip it or its already utiltized and I am getting into data science overrated jargons or if I am right - there is no room for PCA as dimesnionality reduction needs more than 2 or 3 oe even more than 4 fearure to be utlized ....

[@mcanbolat]

You can think PCA as the clustering of variables (features) vs. the clustering of records in Clustering. For example, if you are trying to implement clustering on an employee dataset with Age, Experience, and Salary variables, you will probably not need Age, as it acts similar to Experience (it closely correlates with it). PCA makes that determination and combines those similar variables into one.

[@kbythenewwave7429]

thx!!!

[@NasiimBouha]

did not work

[@alwadi-th6gy]

Thank you very much, Professor. Please, how can I contact you, give me your telegram id