What is a Neural Network?

If you’re reading about artificial intelligence, it’s only a matter of time until you come across terms like “machine learning,” “deep learning,” or “neural networks.” Each of these terms has to do with the way machines learn or operate — the way they learn to capture data and mimic human behavior, making our lives easier and more complex at the same time.

Machine learning is the study of algorithms (the instructions for our devices to do something), and the mechanisms that help machines learn something new. Deep learning is an offshoot of machine learning which uses neural networks to make more accurate predictions and/or decisions. For example, when a shopping website suggests similar items to what you’ve just put in your cart. Neural networks are behind much of the AI we touch every day. In addition to shopping experiences, personal assistants like Alexa or Google Assistant are powered by neural networks.

So, then, what is a neural network? Tough question.

In short, a neural network is the use of millions of artificial “cells” that learn and behave like the neurons in our brains. We have billions of neurons in our heads that transmit nerve impulses. Neural networks are inspired by these human neurons. Similar to what happens in the human brain, a neural network receives information and processes it. But unlike in a human brain, a neural network then produces an output. After a while, a neural network begins to change itself in order to make fewer mistakes. For example, the more we shop for a certain type of item, the more the machine will predict what we may want to purchase based on what the neural network has learned.

Then what does this look like in real life?

A neural network processes fairly small amounts of data at a time, then produces a summary of that data. This technology helps with advanced online searches, personalized shopping experiences, customized marketing, and so much more. Little did we know that we interact with neural networks every single day.