K Means Clustering Intuition

[@shwetadalal1549]

I usually never comment on you tube videos but seeing your videos I am forced to comment. You have got amazing explanation skills and precise knowledge!!!!

[@lemanosmanli2006]

You are amazing teacher, You are my favorite teacher, That With you i can learn Machine Learning. I am continuing to watch you❤ You are the best teacher all over the world 🌍

[@kanhathakur6197]

@Krish, For calculating wcss i saw in some article it is mentioned wcss = (xi - ci)^2.

[@aakashmajumder3064]

yes... the formula which you have written seems to be more appropriate as it gives the distance... addition of the values won't help ...

[@Aldrin32f]

This would be used to decide whether to add more cluster or not also variance can also be used instead of this

[@I_Love_DataScience]

Thank you Krish, thank you thank you 😭😭😭😭😭😭😭 I'm starting to love Data Science because of you 😭😭😭😭😭

[@swapnilsrivastava2039]

Thank you so much sir ... for such an easy explanation .. .. i was so confused about it but now its all crystal clear .

[@manjunath.c2944]

K Means Clustering Practical video is not available kindly provide it sir

[@priyayadav3990]

Yes, it is asking for membership. Please 🙏 sir provide it

[@fredrickosei3138]

good work done ...i enjoyed watching

[@nikhilranjan32]

krish bhai..ur damm aussum. love u

[@vishwajitbhagat9515]

sir should have more subscribers such a good explanation

[@sukanyabag6134]

sir, k means practical implementation tutorial can't be accessed.

[@yuen3660]

hope the first video will be available, cause my exam is coming up in 3 more days x.x

[@user-nk6jt7qi8p]

omg same

[@kaustavmanisarma4481]

Very good explanation

[@sureshkhatri7321]

Thanks for beautiful insight Krish

[@saifullahshafin9094]

Sir, You have lots of video on Machine Learning accept DMM analysis please make a video on it me and my team are waiting for this video.

[@JohnWickea]

love the video sir ❤ great explanation

[@krishnag5734]

Summary of k-means 8:15 and ELBOW METHOD is at 10:50

[@ratulghosh3849]

Thanks for the video sir but the formula of Wccs should be summation of (ci-xi)^2 we need to decrease the dist between the centroid and all the other points in this case...

[@datasciencegyan5145]

the video for k-mean is not present

[@pranavsonawane4]

elbow method 10:50

[@RenormalizedAdvait]

Please make a video on X11, X12, X13 decomposition vs classical decomposition. 🙏

[@akankshasharma9959]

The video you mentioned initially in this video, is not available.

[@ibadkhan481]

simply great 👍🏻

[@dineshkathania2228]

Hi, i appreciate your work , I request to you, can you upload a video of all algorithm with python coding like KNN algo Please. Thanks

[@meetpanchal8809]

Thanks for uploading this,it really helped me

[@vaishnavipandiyan2476]

K Means Clustering Practical : @ its not available for all. it only for members of this channel

[@louerleseigneur4532]

Thanks Krish

[@shadiyapp5552]

Thank you sir ♥️

[@satabdigantayat9852]

here you initialised centroid from the data points only right? i think this is kmeans++ algo. as in kmeans the 1st centroid can be anything i guess (infinity possibility) n not from the data point

[@arunbm123]

nice explanation...

[@dwaipayan10]

Sir, when you are drawing the 2nd line on top of the line joining 2 centroids, is it bisecting and perpendicular to the line joining 2 centroids?

[@DS_AIML]

Great Explanation Krish.So when we have more then one clusters ,we are going to take the average of all the clusters to find the WCSS value

[@jordanstiles5791]

great video

[@vinaykumarreddy9483]

Hi Krish ,thank you for the video.. I appreciate a lot if you could provide the practical video which you mentioned in your video..

[@adiflorense1477]

6:44 Sir, whether the movement of the centroid point is based on the average value we got earlier?

[@itsme-vn8ip]

Yes, we move it to the mean of all the points in its cluster.

[@suvarnadeore8810]

Sir..Please provide K means Clustering Practical Video...Thank you

[@RinkiSingh-ph6oo]

Sir Having one doubt....how to move the point to mean position because it is 2D plot then at what position we will move...pls clear my doubt

[@sandipansarkar9211]

great explanation

[@jayasuryav1674]

Practical video is not available for me how to see? Sir please provide it sir

[@RinkiSingh-ph6oo]

Hello Sir I want to join your channel.....pls tell me the process

[@harshmankodiya9397]

wont the no of clusters also depend on the domain of application also.?

[@E_Gaks]

Really great video krish 😀! I really learnt lots of concepts. I Have a question about wcss score. As it's the distance between each data point and updated centroïd, we shouldn't make a substraction between point of data and centroid point then calculate the mean of these distances according To the number of clusters ?

[@Raja-tt4ll]

Nice Explanation. Thanks.

[@sindhujapateel6455]

Sir can you provide practical intuition for this k means clustring and even in deep learning RNN also

[@riteshgautam1617]

How do you give meaningful label to each cluster. If I do not have a label for cluster then how do we analyse the property of clusters. Please explain

[@krishnaik06]

After clustering u can provide labels and use it for supervised ML techniques

[@DS_AIML]

Basically after K-Means,we will get which row belong to which cluster.Later we can use it like a Supervised machine algo. and get the insight

[@riteshgautam1617]

@@krishnaik06 Hi actually my main question is how do i provide a label to it. If i have 1000 customers and the algorithm has divided into 4 clusters then how do i provide meaningful label to each of these cluster. Do i manually look at the records, analyse and then give a label or there is any technique to figure out the label for each cluster

[@pratimadhar9202]

please share K Means Clustering practical video.

[@arungovindarajan93]

Hi Krish, Initializaton of k-centroid points will have sensitivity problem ?? That means if we choose the randomly k centroid points initially we might not be sure that we get desirable cluster. Then how to get rid of this problem?? Please explain

[@krishnaik06]

There are some parameters that we must select while initializing the KMeans library

[@siddharthasharma9316]

Density based clustering

[@sakshamdutta6366]

How do you define abrupt decrease? Does it mean differentiation of wcc with respect to k must be highest?

[@21_jadhav_rajendra84]

I think we can take the Y axis as reference when considering sudden decrease

[@bhushandhuri2660]

Does anyone know how k-means maths works when there are 4 features? Also if you know of any resource available on the internet then please comment it down.

[@swapnilsrivastava2039]

Sir k means practical video pls ...it is not available.

[@abhivdos123]

K means Clustering Practical vdo has been made private, Can you please make it available

[@subhamsubimanbiswasrevenue7891]

In Categorical data , kmeans is not applicable, and one hot encoding is wrong for clustering.

[@brahmanandakabi8796]

If there is a noise what happen

[@snigdharay8847]

Here when we divide the data into two clusters at first then also we are getting those groups by setting the centriod of our own. they why you are again taking its average and dividing it. That is confusing.. Any body reply@

[@prashanthpandu2829]

To get the appropriate group or cluster of datapoints that are having similar features.

[@upendrakumarpanda9009]

Guys, please correct me if I am wrong... While choosing k-value for Classification : Should we need to take only odd numbers between 1-40 ? because even value of K could leads to a problem in KNN..

[@anubhavgupta8146]

Yes, u are absolutely right.

[@parthnigam1782]

Yes it's leads to over fitting

[@vaaidooyt1837]

@Krish Naik #KrishNaik Sir , PRACTICAL IMPLEMENTATION IS SHOWING AS " MEMBERS ONLY CONTENT", Can you please make it public

[@sarankumars1982]

I still think the elbow method should be explained in detail

[@jetendramulinti6443]

Thanks Sir

[@aniketthorat3957]

Practical video for k means Is private plz make it as public

[@sarahelariane9355]

do we need to split data to train and test in Kmeans to overcome the overfitting or no overfitting in unsupervised learning

[@priyanksisodia5889]

No ,because Kmeans is a unsupervised algorithm ,which means we do not have labels ,so the point of splitting the data is wrong

[@Sunilgayakawad]

Op sir🔥

[@srinivashini2867]

👍👍

[@adiMallya]

videos are made PRIVATE, and shows UNAVAILABLE

[@atharvashirode9711]

@krish naik can u provide thus?

[@Ashuraj373]

first video is private please make it public

[@NileshPatil-li1so]

Hi Krish i tried to purchase have got 900 deducted from my account

[@NileshPatil-li1so]

9000 rs

[@NileshPatil-li1so]

pls help

[@sunilsurya6305]

K Means Clustering Practical video is not available kindly provide it sir