In the context of AI, a Neural Network is the construction of artificial neurons. What this means is that different pieces of placeholders (with complex mathematical formulas) are organised in a way that is inspired by how the human brain would form relationships or connections between multiple bits of information.

Some recommended reading before we dive into this topic is our other posts defining AI & Machine Learning.

Neural networks are a very efficient way for machines to navigate large quantities of data. After all, the human brain is the perfect inspiration of how to very quickly form a conclusion based on the data it is presented with.

A neural network works by mapping out and linking elements of data (show-offs call them perceptrons). A very straightforward example of a neural network is explained in this below:

This neural network is constructed to give insight into the most popular item being purchased and from where on an e-commerce website. Normally, you don’t need a neural network for this but for the purpose of simplicity we will keep it confined to the example here:

In relation to the above illustration, the blue circles would be information extracted from the e-commerce website.

Each blue circle represents:

• The number of items purchased per location.

The purple circles represent the neural network forming relationships to find patterns between the data and orange is the output.

If we apply this to a use case this neural network would be able to see that shoes were most popular in an exact location. If they wanted to see where T-Shirts were purchased the most, they would also be able to see that with modified configurations. The orange circle could be any result the site owner wants it to be based on the instruction given to the machine using the neural network.

If you have any questions about AI we would love to hear from you. Contact us by e-mailing info@speire.ie.