Support Vector Machines: A Visual Explanation with Sample Python Code

Показать описание

SVMs are a popular classification technique used in data science and machine learning.

In this video, I walk through how support vector machines work in a visual way, and then go step by step through how to write a Python script to use SVMs to classify muffin and cupcake recipes.

In Part 1a, I visually define the following terms:
- Margin
- Support vectors
- Hyperplane

In Part 1b, I go through the following steps in a Jupyter Notebook:
- Import libraries (pandas, numpy, sklearn, matplotlib)
- Import data
- Prepare the data
- Fit the model
- Visualize results
- Predict a new case

In Part 2, I talk about ways to tune the model:
- Higher dimensions
- Multiple classes
- C parameter
- Kernel trick (RBF with gamma)

In Part 3, I talk about the pros and cons of SVM.

Рекомендации по теме

Комментарии

That was the best tutorial I have seen on SVMs. Impressive. The slides were very well done and the explanation was simple. It's nice to see an tutorial without the math. When you start to explain the math it gets very complicated and it is easy to get lost.

timharris

Your explanation is just the best, simple and clear; please make more tutorials and cover other topics, and thank you so much

hadjdaoudmomo

Sometimes I think when sitting in class, "I must just not be that smart and these concepts are just too intense to comprehend". And then youtubers, such as yourself, make me feel much better about the same topic, in less time than sitting in class. Thanks so much for taking the time to make videos such as this to help students around the world. This stuff is so cool when it makes sense!

GAment_

This is the best tutorial I have ever watched on SVM. She explained everything in a simple way. It was very easy for me to understand SVM. Thanks, Alice!

reazahmed

Cup cakes to SVM depth. Amazingly genius idea . Thanks for explaining it so clear!

shekharkumar

Alice you have the natural talent to teach, very impressive. I am waiting for your next video !

AvinashReddy

Probably THE simplest way of explaining SVM..I'm glad that I stumbled upon your tutorial..thank you!

sandeepahuja

The first three minutes of this video helped me develop a clearer understanding of SVMs than numerous previous attempts, making clear just how simple the underlying idea is, and immediately clarifying the jargon straight up front by mapping to simple 2D language.

Anyone that can provide such a simple and mentally sticky image where many others have tried and failed has got to be worth subscribing to. Really well done. I can't wait to watch your other stuff.

jimwest

you just saved me, I was trying to understand SVM the whole day and now I get it, your explanations and examples are great

vincentdepassier

Points that I cont understand for months, this tutorial taught me -> Widest ROAD. Thank you

asimsultan

Positive points Alice commented about 2 classes and also multiple classes. Explanation about 2D, 3D and more dimensions data were very well done. Great explanation, thank you for sharing with us.

vivianacoccomariani

Just blew my mind, I've been normalizing data since I started my online data science bootcamp, but haven't really understood the concept, the first 5 minutes of your video was one of those shining light moments! Thank you Alice.

martinrussell

Really its is the best tutorials on SVMs. Brief and sequential with focus on application oriented learning. Great.

technoreviews

This video is one of the best ML videos I've ever seen! I love how it all started with a random curiosity about classifying muffins and cupcakes, then you show us step-by-step how you actually solved this problem. Thanks so much for this tutorial.

cbrtdgh

Best tutorial for SVM as per now what I saw other youtube videos on SVM. It includes the best way of representation using actual code and doing analysis. Keep It up.

NxtGenTech

Your video is still actual, even in 2021! Thank you a lot. I would be happy to see your videos about other methods!

ndhpwgd

Honestly, the best video explanation I have seen so far on SVM.

xZzzit

one of the best lectures I've listened to regarding SVMs.

eranozsarfati

This is one of the best programming/ data science tutorials I've ever seen. You're a natural teacher. I see you haven't made any videos in some time, but if you ever start teaching again know that you're changing lives!! For instance mine :)

dreamphoenix

I bought a course on ML and had to resort to youtube to understand SVM since they got very technical. This is by far the best explanation I have seen!

utkarshjain

Support Vector Machines: A Visual Explanation with Sample Python Code

Support Vector Machine (SVM) in 2 minutes

Support Vector Machines: A Visual Explanation with Sample Python Code

The Kernel Trick in Support Vector Machine (SVM)

Support Vector Machine (SVM) in 7 minutes - Fun Machine Learning

Support Vector Machines Part 1 (of 3): Main Ideas!!!

Support Vector Machines: All you need to know!

Support Vector Machines : Data Science Concepts

Support Vector Machines - THE MATH YOU SHOULD KNOW

68 - Quick introduction to Support Vector Machines (SVM)

Support Vector Machines: Visual Explanation w/ Practical Demonstration

Support Vector Machine - How Support Vector Machine Works | SVM In Machine Learning | Simplilearn

Support Vector Machines (SVM) - the basics | simply explained

How SVM (Support Vector Machine) algorithm works

Support Vector Machine (SVM) Basic Intuition- Part 1| Machine Learning

16. Learning: Support Vector Machines

Support Vector Machine | Face Detection

Visualizing Support Vector Machines: Unraveling the SVM Algorithm

Support Vector Machine - Georgia Tech - Machine Learning

Support Vector Machines in Python from Start to Finish.

Support vector machine: XOR example

Machine Learning Tutorial Python - 10 Support Vector Machine (SVM)

Python: Support Vector Machine for Classification (Tutorial)

Support Vector Machines (SVMs): A friendly introduction

SVM Algorithm in Machine Learning | Support Vector Machine | Scikit Learn Tutorial | Intellipaat