Support Vector Machines — Lecture series — Part 1: An introduction to support vector machines
NOTE FROM THE AUTHOR TO THE READER
For many, the field of machine learning and artificial intelligence has been shrouded in mystery and even though there are many people who claim to understand how the various machine learning algorithms work, there are actually very few people who appreciate and deeply understand the beauty within the mathematical ideas on which these algorithms are built on.
In this lecture series, I will be sharing my knowledge on the theoretical and mathematical ideas behind various machine learning algorithms. I will try my best to present the ideas I will be teaching in a very straight-forward and simple way so that you will derive the most amount of benefit from these lecture series.
It is my aim to transform you into a machine learning expert who has a deep understanding for the field and can possibly contribute to it by devising your own learning algorithms.
You do not need to be a machine learning enthusiast to benefit from this series. Heck, you do not even need to be a computer scientist. My aim is to make this accessible and understandable to everyone who has at least a basic education in english and mathematics.
Let’s begin :)
The learning objective
My main objective here is to enable you to have an intuitive understanding of the concept of support vector machines and what they are useful for. It is quite a simple and beautiful concept.
The main question
Consider the image in Fig. 1 below:
In the image above, we see 2 different kinds of objects. One object is of the form X and the other object is of the form O.
We also see that all of the X objects are clustered in one area whiles the O objects are clustered in another area.
Now the main question is, this how can we come up with the best possible line to separate the two different clustered areas?
Consider the image in Fig. 2 below, which line best separates the two clustered areas? Line 1 or Line 2?
Determining the line will best separate the 2 different clustered areas of objects in the image above is the awesome job of a support vector machine.
Hence support vector machines are machine learning algorithms that can be used to formulate the best possible plane (that is, line in this example) that can separate some given clustered areas of example objects.
In the next lesson, we would be talking about how vectors :)