Introduction to K-means - Choosing number of clusters

[@TheUmaragu]

Nice explanation; I loved the way you used the two extreme cases for clustering.

[@FrOsTyBeArKiD]

This was very very clear. I enjoy the examples showing the extreme scenarios to make the optimum example hit home.

[@alexandrsgrind3743]

even if my ears hurt by your hardcore indian accent, your way to explain is brilliant, sir!

[@akhippal]

This is really good. Thank you !

[@jorlop97]

Very useful and helpful. Thanks

[@easyml1234]

Thank you for your kind words. Please help spread the word 🙏

[@dawitkim6650]

It's a very short video but it's much helpful.

[@Dave-px2xc]

Can you go in depth to how to compute the WSS or WCSS (Within Cluster Sum of Squares)? Very helpful video indeed.

[@easyml1234]

Thank you for your feedback will find a way to create a video that explains the calculation of WSS.

[@farsikogama6114]

@@easyml1234 what is the video to calculate the WSS?

[@easyml1234]

@@farsikogama6114
I am sorry I am busy with some personal tasks so I couldn't get time to make the video but the R CODE FOR WSSPLOT FUNCTION CAN BE FOUND HERE :-
rpubs.com/violetgirl/201598
I have however created a video where I have explained how to deploy the WSSPLOT function :- ruclips.net/video/KmYUE7Of5rU/видео.html

[@sanihunsamro2551]

this is helpfull, thanks!

[@karthika9592]

Very nice presentation, Clear explanation! What can we do if no such dip formed in the graph while finding optimum no of clusters

[@easyml1234]

Hi thank you for your comment. If there is no clear 'kink' then you can experiment with various values. For instance if the graph is gradually sloping around number of clusters 2,3 and 4 then input each of those values to see which produces distinct clusters. TBH in real life data very rarely will we be able to choose clusters easily🙈

[@bestofthebest2644]

Great teaching

[@easyml1234]

Thank you :)

[@vinijcobatgamildotcom]

can you make video on logic regression.

[@easyml1234]

I will definitely try and incorporate that :)