Machine Learning Algorithm | Understand KNN Algorithm in 5 Minutes

wangerxian

Machine Learning Algorithm | Understand KNN Algorithm in 5 Minutes [Copy link]

K-Nearest Neighbor Algorithm | K-Nearest Neighbor Algorithm KNN explained in an intuitive way
---
What is K-Nearest Neighbor Algorithm?
K-Nearest Neighbor, also known as K-Nearest Neighbor Algorithm, is a learning method that allows computers to imitate our daily decision-making. It is example-based and very intuitive, just like when we choose restaurants, clothes or travel attractions.
---
How to understand this algorithm?
Imagine that you are in a new city and want to find a good restaurant. What would you do? You might look around to see which restaurants have more customers, or ask local residents.
This is the idea of K-Nearest Neighbor:
- If all friends around you recommend a place, you might like it too.
- If they all say it is not good, you may not go there either.
---
What is the 'K' in K-Nearest Neighbor?
"K" is actually very simple, it stands for "several" neighbors.
- K=3 Consider the suggestions of the three closest friends.
- K=1 Only listen to the closest friend.
- K=n Consult the opinions of all friends.
In short, K-Nearest Neighbor is a method of making decisions by considering the opinions of people around you.
---

See the picture for details, and there is also actual code! I hope this article can help you quickly understand the K nearest neighbor algorithm!

freebsder

Clear and concise, 5 minutes is enough.

chejm

The machine learning KNN algorithm shared by the host is detailed, illustrated, and easy to understand, which is very helpful for understanding the KNN algorithm.

hellokitty_bean

From this perspective, KNN is indeed a lazy algorithm... Haha

However, how many K should be given? Is it a random decision or is there any basis for it?

wangerxian

hellokitty_bean posted on 2024-4-25 08:58 From this point of view, KNN is indeed a lazy algorithm. . . . . Haha, but how many K should be given? Is it a random decision or is there something else...

Then let the algorithm test K and get the best result.

wangerxian

chejm posted on 2024-4-24 21:44 The machine learning KNN algorithm shared by the host is detailed, illustrated, easy to understand, and very helpful for understanding the KNN algorithm

Yes, if I find a good article, I will share it with you.

wangerxian

freebsder posted on 2024-4-24 20:53 Clear and concise, 5 minutes is enough.

I will share more of this kind of graphic algorithm explanations in the future.

freebsder

wangerxian posted on 2024-4-25 09:55 I will share more of this kind of graphic algorithm explanation in the future.

Haha, thanks for sharing. The island country has done a great job in this area, comics xxx and so on.

wangerxian

freebsder posted on 2024-4-25 17:21 Haha, thank you for sharing. The island country has done a good job in this area, comics xxx and so on.

China is also doing very well now, especially 3D animation~

freebsder

wangerxian posted on 2024-4-25 17:59 China is also doing very well now, especially 3D animation~

There are indeed quite a few such resources on Bilibili and Tik Tok now, but they feel very scattered.