K-Means Clustering From Scratch in Python (Mathematical)

In this video we implement K-Means clustering from scratch. We look at the theory and the mathematics behind it and then we use NumPy to put it into code.
◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾
📚 Programming Books & Merch 📚
🐍 The Python Bible Book: www.neuralnine.com/books/
💻 The Algorithm Bible Book: www.neuralnine.com/books/
👕 Programming Merch: www.neuralnine.com/shop
💼 Services 💼
💻 Freelancing & Tutoring: www.neuralnine.com/services
🌐 Social Media & Contact 🌐
📱 Website: www.neuralnine.com/
📷 Instagram: / neuralnine
🐦 Twitter: / neuralnine
🤵 LinkedIn: / neuralnine
📁 GitHub: github.com/NeuralNine
🎙 Discord: / discord
Timestamps:
(0:00) Intro
(0:17) Theory & Math Background
(7:21) Coding K-Means From Scratch
(32:37) Outro

[@pitaeata8493]

this is great, thank you. it feels good to understand something and be a little closer to understanding machine learning or how to use it properly.

[@AlexandLupand]

I'm glad I found this tutorial!

[@linusolmin9313]

Thank you very much!

[@Hickssama]

Thank you sir

[@tcgvsocg1458]

really interesting

[@SeziManuhutu]

your explanation is great. but i want to ask, what if we want to know what data is move to different cluster. like at first the data is in 0 cluster but then after that it's move to the other group. how do we know the number of data that is move to another cluster?

[@Larzsolice]

I take random points from my data as initial centroids, less computations since you only need the set a random integers for indices.

[@bibekmaharjan8969]

Can you share the code for it

[@Larzsolice]

@bibekmaharjan8969 not sure how here, but the Python random package has a random.choice() or something like that. You can use the data points as the iterable and choose k of them randomly like that.

[@bibekmaharjan8969]

@@Larzsolice alright. Thanks a ton

[@bijayamanandhar3890]

It's a great tutorial. Beside everything, I just didn't understand why and how it was assumed to have 3 centroids for the example dataset where as you assumed the dataset has no label (unsupervised). Appreciate if you can elaborate. Thanks,

[@NeuralNine]

The number of centroids is the same as the number of clusters. You can either just pick one or you can use techniques to find the "optimal" or at least a good number of clusters. I have a video about this on my channel.

[@Ali110.69]

Hello, I had a question, I wanted to know if I want to work in machine learning or data science, I must use Anaconda and Jumper Lab or can I use Pycharm? I asked this question to many people and they said I should work in Jupyter Lab I can't get along with its environment. I can work inside pycharm or vscode. Your answer can help me a lot. Thank you

[@user-ti6zf5mj4x]

Better use VS code to have a JN architecture, and if it creates issues, you may use Pycharm!

[@sekiro_19]

use Jupiter notebook

[@philtoa334]

Thx_.

[@user-gk5bz6vd9j]

hi i am getting this error can you tell how to solve it
ValueError: 'c' argument has 200 elements, which is inconsistent with 'x' and 'y' with size 100.

[@aravindputtapaka5147]

I want a python code to convert handwritten image into plain text with accurate i have tried buti didnt got you can try it and show it me sir and plz respond to this comment bcz i am searching for this very curiosly...

[@AnmolKumar-so8lh]

use OCR(optical charcter recognation) for that using ocr you can extract text from the image