How SVM algorithms work

Written by Arun Krishnaswamy

This article is a part of the series to explore different ML algorithms and simplify them. The goal of the series to help common people understand algorithms. When someone comes across “SVM” — they will know what it is!

Part 1 — SVM


AI SVM Machine Leaarning

2 colors of balls on a table to be separated

Let us Put a Stick


AI SVM Machine Learning

Put a Stick to Separate

What if someone adds more balls to the table

AI Data Machine Learning SVM

More balls added to the table

SVM solves this problem as follows — Puts a stick on the table by having a big gap on either side of the stick as possible.

AI Machine Learning Data SVM

SVM puts a stick by having a BIG GAP between the Sticks and the balls on either side

What if u don’t have a stick — can SVM still separate?

FLIP the table –Throwing the balls into the air. Put a sheet of paper and slip it between the balls.

AI SVM Machine Learning

Flipping the table and Inserting a Paper between the balls

The balls are split by a CURVE

AI SVM Machine Learning

Balls split by a CURVE instead of the stick — but they are SEPARATED

SVM Algorithm Glossary

Balls — data,

The stick — classifier,

The biggest gap trick — optimization,

Call flipping the table — kernelling,

Piece of paper— a hyperplane.

Stay tuned for more demystification…..

About the Author

Arun volunteered for Humans For AI, a non-profit focused on building a more diverse workforce for the future leveraging AI technologies. Learn more about us and join us as we embark on this journey to make a difference!