Which is probably more accurately called Machine Learning. It is new breed of science which has been around a long time but only now has the technology caught up. It is talked about a lot both the good side and the less than good application but whatever people’s views it is here to stay. All I want to do is give you a simple, beginners guide to understanding of what this is and how to at least get started.
The brain has billions of them and they allow the brain to function as it does. What Artificial Intelligence (or Machine Learning) tries to do is emulate it by using something called Artificial Neural Networks. They don’t work in exactly the same way but do try to emulate the process of connecting them and learning.
A single Perceptron
Click on the button to see one in action
Click the mouse to see it learn
This is a simple AI using one perceptron.
The program is training itself to work out what circles are above the line and which are below.
It doesn’t know and has to learn by making incremental adjustments.
It makes a guess
The white circles should be above the line and the black circles should be below the line.
Every time you click you are training it against what should be the output.
It may take one training cycle or several.
It all depends where it starts from in its initial guess.
A Neural Network
A simulation of a neural network. This recognises handwritten numbers