From Scratch: How to Code K-Means in Python (No Sklearn) for Machine Learning Interviews!

K-means Algorithm Step by Step in Python (No Sklearn) | Data Science Interviews | Machine Learning Interviews

🟢Get all my free data science interview resources

// Comment
Got any questions? Something to add?
Write a comment below to chat.

// Let's connect on LinkedIn:

====================
Contents of this video:
====================
00:00 Topic of this video
00:52 K-means Overview
03:05 Main and helper functions
03:51 Implementation
08:58 Time and Space Complexity

[anudeepandy]

The way you explain is simple but magical. That's an unique style. You explain concepts to the point and only the necessary and important points. Please demonstrate the shortcomings and pitfalls of the algorithms too. Like when the algorithm will fail and when is it not a good choice. Also if you can explain data structures and algorithms that would be very helpful! Keep posting we support you! #YouAreAwesome

[徐宁远]

Emma 老师讲的太好了，尤其喜欢你对coding part的讲解，非常详细，清楚，concise。

[jiahuili]

These details really help understand the algorism on another level. Thanks a lot.

[hrushikeshshelar]

Great Video. Please keep posting such videos. Explanation was very well and to the point.
Looking forward to more such videos.

[williammcgrath]

Very helpful. Showing the math and converting it into code side-by-side was especially helpful. I look forward to seeing more similar videos!

[JYphotography]

Awesome explanation. I am more clear about each step after rewriting each step on my own.

[jaybhanushali]

I tried your code, many times it gives Zero division error because of centroid initialization... any solution for that?

[shirleywang]

it's very clear! thank you very much. the way you wrote python is far more structured and clearer than me. I would also want to know how you interact with the interviewer and how you actually code during an interview, rather than explain the algo and code using PPT. (i know that's a lot to ask, but if you could show one or two cases, it would be super helpful!

[pingdeng]

This is a real nice Video, have detailed introduce the k-mean cluster using simple way, looking for more videos!

[guogaibnu]

Awesome video! Super clear, yet concise.

[ryangeezy]

Super clear explanation. Thanks for posting!

[TanmayShrivastava]

This was really great and crisp clear. I will look forward to the one you will do on logistic regression, as it might just be the most basic classification algorithm. Thanks for your efforts.

[nhandam]

Brilliant content. Very clear explanation. Thank you very much!

[TL-tcjq]

Thank you so much! Very clear and concise explanation. This is very helpful!

[bijayamanandhar]

Thank you so much for such a beautiful explanation on KMeans Algorithm. I was looking for an answer to the interview question "How do you determine how many clusters to choose for a most accurate result by using the unsupervised machine learning KMeans Clustering algorithm?". I hope you will surely help me. Thanks,

[dineshbuswala]

Awesome, just want to highlight one point if the dimension of training is higher then time complexity would be O(n*k*d*i)

[YEIYEAH]

Outstanding. Thank you for the explanation.

[dallalstreet]

thank you Emma!! superbb video . Please do video on SVM also and other algorithms same way

[jianzezhou]

Haha when I heard that you gonna analyze the time complexity I thought you gonna calculate the number of iteration in terms of threshold. But anyway, very comprehensive video, thanks for explaining with clarity and details.

[arrowfunction]

thank you so much. i need k means python code without libraries and using real data. do you have any recourse for it?