Silhouette Score for clustering Explained | Silhouette (clustering)- Validating Clustering Models

[@shaneglean217]

The explanation is so concise and simple, even in spite of the fact that english isn't your 1st language. Amazing work my dude.

[@aaryalad7775]

I'm glad i found a teacher like you

[@ravisavaliya4333]

at first it was very difficult for me to understand this, but you explained this in a very easy way....THANKYOU AMAN SIR FOR YOUR EASY AND INTUITIVE EXPLANATION!!!

[@UnfoldDataScience]

It's my pleasure

[@MyerNore]

Clearer than my professor, for sure

[@Pannafreestyle]

Sir you are the best teacher for DS ! Keep up the great work

[@isaacyekoniya8067]

This is simple and well explained, thank you Aman!

[@UnfoldDataScience]

Most welcome!

[@victorcardenas9621]

Simple but powerful, thanks and greetings from Colombia!

[@ankitsapkota3713]

Never expected the conceptual delivery of this level without random steps out of the air in this accent.

[@JuliaCh47]

Thank you so much, best explanation I've seen on the Internet!

[@UnfoldDataScience]

Thanks a lot for watching.

[@saijaswanth3036]

Thank you for clear explanation

[@RezoanurRahman]

You explained it so well even a 5th grader would get it. Its so easily explained!.

[@RoopaNandu-k8b]

Aman, your explanation is really good. Very precise and in detail

[@uchennanwosu5327]

Excellent illustration and explanation.

[@kushik.naveen]

This is so useful for understanding & revising. Amazing work! Thankyou!

[@ShaliniK-s5m]

Ur Explanation is simple and understandable,TQ sir

[@lhadz7290]

Thank you so much for Sir, I am so glad I came across your video. I look forward to learn more from you. Thank you again.

[@UnfoldDataScience]

So nice of you

[@Anon11229]

Too good...Way better than online courses that take thousands and thousands of money keep up the hard work

[@UnfoldDataScience]

Thanks Zaid.

[@Felipe90820]

Amazing explanation! Congrats and thank you!

[@mdyounusahamed6668]

very concise and simple explanation.

[@paulholzapfel1856]

Perfect explanation. Thank you Aman

[@UnfoldDataScience]

Welcome Paul. Your feedback is precious to me.

[@nandankakadiya1494]

Amazing explaination thank you sir

[@UnfoldDataScience]

Welcome Nandan.

[@Suno-99]

beautiful explanation, keep it up

[@sangeethaagoudar2175]

Thank You so much for this video sir

[@awd_junkie]

Clear and right to the point! Thank you for your work! Much appreciated!

[@UnfoldDataScience]

Thanks for watching Chris. Stay safe. Tc

[@Krishna-pm8ty]

Excellent Amen. Perfect explanation👏👏👏

[@nandinik7816]

Always to the point...and you dont waste any time actually...

[@UnfoldDataScience]

Thanks again.

[@KHakimin]

This is gold! thanks for putting it very simple to understand

[@UnfoldDataScience]

Glad it was helpful Khairul.

[@shaheerrehman3151]

Thank you. Much love to you from Pakistan

[@UnfoldDataScience]

Thanks Shaheer, pls share with friends if you like the content.

[@izb1275]

Great explanation !

[@amoghbharadwaj9252]

beautifully explained!!! Thank u so much

[@UnfoldDataScience]

Thanks Amogh. Happy Holi.

[@puodzius]

Very good explanation, thanks

[@UnfoldDataScience]

Welcome.

[@nicolaevasiliu9430]

Amazing explanations! THANK YOU !!! You are an excellent teacher !!!

[@samiayasmin5782]

Thank you so much. Simple and clean explanation.

[@liqs_xd]

Great video, thanks for your time!

[@UnfoldDataScience]

My pleasure!

[@sanjithaamarathunga9949]

Thank you! Simple and well explained!

[@UnfoldDataScience]

Glad it was helpful!

[@dosendaring]

This is very good explanation, thank you Sir....

[@UnfoldDataScience]

Welcome Yudy.

[@shaikhuzma786]

Tqsm sir so well explained ❤

[@the.quantum.paradox]

Very simple and easy to understand!

[@sourabhpramanik3957]

Its really a simple and good explanation for anybody to understand...thanks for making this simple explanation video.

[@UnfoldDataScience]

Welcome Sourabh

[@dbapple]

thank you, straight to the point!

[@YouTribalChief]

What a gem of an explanation you give bhai! Truly you make things so simple and understandable! You deserve more!

[@UnfoldDataScience]

Thanks a lot. Pls share with friends as well. Have a nice weekend ahead.

[@parvsharma8767]

Thanks bro for explaining so perfectly

[@UnfoldDataScience]

Welcome Parv.

[@omniscienceisdead8837]

you are the best bro!

[@vallimuthaiyah5098]

Thank you sir for very clear explanation with example.. 👍

[@UnfoldDataScience]

Welcome Valli :)

[@blackyogurt]

Great Explanation . thanks

[@sajjaddehghani3280]

Thanks for the very simple and easy to understand explanation

[@UnfoldDataScience]

Welcome

[@venkataprashanth9831]

great and simple explanation, a big clap to ur efforts

[@UnfoldDataScience]

Thanks Venkat.

[@keinestudiere-chillzone633]

Straight to the point. Thank you

[@UnfoldDataScience]

Thanks a lot for watching

[@MohdRashid-hi7cp]

Very nice explanation

[@UnfoldDataScience]

Thanks Rashid.pls share with friends also

[@anshuraghav5620]

Great video

[@muhammedthayyib9202]

Clearly explained!

[@vamsikunchala]

Amazing!

[@legolife3012]

Thanks a lot for this simple explenation

[@UnfoldDataScience]

You are welcome

[@akashpb4044]

Awesome ...plain and simple 👍🏼

[@UnfoldDataScience]

Glad to hear that Akash 🙂, please share with others as well who could be benefited from such content.

[@kavyasharma4738]

how can i thank you brother . you explained this topic so easily.

[@kamal_douma]

very very good explnation

[@beprepare8128]

Very well explained sir

[@UnfoldDataScience]

Thanks and welcome

[@diyap7343]

This is amazing!

[@UnfoldDataScience]

Thanks Diya.

[@anisadillap.7131]

oke i understand, thankyou sir ...

[@UnfoldDataScience]

Welcome

[@PranavVashisth]

verry well explained !!!!

[@UnfoldDataScience]

Thanks Pranav.

[@theone-jz2rf]

keep up the good work!!!

[@UnfoldDataScience]

Thank you.

[@optimalsolutionprogramming1299]

Thank you for your help. Really great explanation!

[@UnfoldDataScience]

Welcome

[@mohe4ever514]

Very well explained!

[@UnfoldDataScience]

Glad it was helpful!

[@veyselaytekin8734]

thank you

[@bipulnath4602]

Amazing explaination..plz make a video on Dunn index.

[@UnfoldDataScience]

Thanks Bipul.

[@bangarrajumuppidu8354]

excellent

[@UnfoldDataScience]

Thanks a lot.

[@gnaneshgn8341]

Nice video Sir thank you. please make an video of EDA sir ..!

[@UnfoldDataScience]

Thanks Ganesh.

[@stevemungai3542]

Crystal clear

[@SACHINKUMAR-px8kq]

Thanks you so much sir

[@UnfoldDataScience]

Welcome Sachin.

[@jeanhwang18]

Thank you so much for your explaination. I have a small question: after I got point i ' s silhouette coefficient, I calculate other two points in cluster A, then I take the average of the three silhouette coefficients, let's say it's NO.1. What we can learn from No.1? If No.1 is close to +1, it's good, it means data points in cluster A are most similar to each other, right?

[@neajinnovations8587]

I got all the concept clear, but please when is it appropriate to use silhouette distance? it when you have mixed data? or mix data is best for elbow method?

[@Live_life_with_no_regrets]

is there any way to run it in SPSS? Is it correct if I use the average of the Euclidean distance of the cluster?

[@vishwass5103]

Hi Aman thanks for this explanation. Please explain about assumptions of k means clustering

[@UnfoldDataScience]

Assumption at high level is, you data has clusters and centriods.

[@dr.zliiiiii]

nice video

[@ranajaydas8906]

Sir please make a video on PCA . It would be very helpful

[@UnfoldDataScience]

Sure Ranajay.

[@sandipansarkar9211]

finished watching

[@chrisx742]

Thank you.

[@SurajReddy-hx2dx]

concept is well explained . But what do we do after this step. You have only calculated the silhouette score for one point. Do we need to calculate for all the points? What is the range of values present for silhouette score? When is it good or bad?

[@mahipatil369]

hello sir.. great video i understand this topic very well,, thank you so much... just want to ask did you make video of implementation of silhoutte score in python?? if yes plz provide link..

[@UnfoldDataScience]

Thanks a lot. Its very simple to do in python - see this link - stackoverflow.com/questions/59919627/how-to-calculate-the-silhouette-score-for-each-cluster-separately-in-python

[@sampathvinaykumarreddymajj790]

Topic is very well explained sir .. But should we learn DB Index, Dunn Index, Jaccard score ... or only this Silhouette score is enough

[@UnfoldDataScience]

Thanks Sampath.

[@himanshumangoli6708]

As we have find cohesion and separation for only one point in one cluster A, so we have to find Silhouette score for all points for all points in same cluster so just to ensure that points are correctly clustered ?? Please explain??

[@UnfoldDataScience]

Yes , one point is just for example.

[@megalaramu]

Hi Aman, I do have a general question in clustering -I have heard of people do clustering in the dataset initially and then train separate model for each cluster. my question is lets take an example of telecom churn - there will be many services like internet, Phone, Dish etc. is there anyway we can cluster like internet users in one cluster, phone service in one cluster like that.

[@mmmommm237]

Great, ty!

[@prashanths4455]

super bro

[@UnfoldDataScience]

Thanks Prashanth

[@arisurya9867]

Nice video sir. But i have problem when calculate cohesion in cluster just has one data point. How i should set value of cohesion? 0 or 1?

[@UnfoldDataScience]

Thats not a good clustering process, clusters need to be reformed.

[@arisurya9867]

Thank you sir 🙏

[@subhasisdutta3711]

you have said "Min distance of A1,A2,A3 is the value of b ". Here you are talking about mean value of three or the which one has the minimum value among three , we'll chose it. Please clear it once, if possible

[@UnfoldDataScience]

HI Subhasis, A1, A2 etc are "mean" Distance from various other clusters ok. Now let's take att these A1 TO AN values and take the minimum from this set. This mimum is b.

[@subhasisdutta3711]

@@UnfoldDataScience thanq for the clarification. Have a blessed life ahead

[@gretchenbekker2371]

How do I get the amount of data points within the clusters ?

[@UnfoldDataScience]

Levels can be found out after clustering then we can count using pandas.

[@pruthvi491]

How to get silhouette score for spectral clustering

[@UnfoldDataScience]

I Will check.

[@xendu-d9v]

thanks sir

[@UnfoldDataScience]

Most welcome

[@parikshitgurjar5545]

hello sir, Plese make a video on the explanation and implementation of ppf,pdf,cdf in python with simple explanation and understanding, I am very confused in this.

[@UnfoldDataScience]

pdf cdf video is available Parikshit. Please search in my channel.

[@parikshitgurjar5545]

@@UnfoldDataScience Thankyou sir❤️

[@handsomemehdi3445]

Thank u for simpler explanation! A question: In our case, it seems that for both clusters, b should be the same unless we have more than two clusters, (or since those clusters are mutually nearest) doesn't it?

[@lkjhlkjh1954]

I think that’s why we do not calculate silhouettes score for less than 2 cluster

[@Chaseaction-outdoor-activities]

thank you!!!

[@UnfoldDataScience]

Welcome.

[@nerdymath6]

is 0.4 is good silhouette score?

[@UnfoldDataScience]

Dont think so

[@christopherlegarda5164]

This guy is god!

[@salilshukla08]

thanks

[@UnfoldDataScience]

Welcome.

[@TheGentlemanGR]

So which score is better? -0.2 or 0.21 ?

[@dariomelconian9502]

Is 0.39 a bad score?

[@UnfoldDataScience]

can be better

[@iftikhar58]

clear

[@UnfoldDataScience]

Thank you

[@abeerhamid]

How can i read the graph of silhouette score? Most important thing you didn't tell

[@UnfoldDataScience]

Positive Scores (close to 1): Samples are well-clustered, and there is good separation between clusters. Scores around 0: Samples are on or near the boundary between clusters, indicating potential overlap. Negative Scores: Samples are likely in the wrong cluster.

[@sameeransari9129]

apki english samajhne me time lg jata hai.. ki aap kya bolna chahte ho