Deep Learning Fundamentals - Classic Edition

What is deep learning?

In this post, we will be answering the question, what is deep learning?

This entire series will be dedicated to covering many topics within the deep learning field, and it will take many posts to fully explain these subjects, their applications, and their technical implementation.

Additionally, these posts are placed in a particular sequence within this deep learning series. Certain concepts covered will build on concepts that may have been previously discussed, so make sure to check out previous posts within the series if you're not already familiar with some of the terms or examples we're using.

Let's give a definition for deep learning.

With deep learning, we're still talking about algorithms that learn from data just like we discussed in the last post on machine learning. However, now the algorithms or models that do this learning are based loosely on the structure and function of the brain's neural networks.

The neural networks that we use in deep learning aren't actual biological neural networks though. They simply share some characteristics with biological neural networks, and for this reason, we call them artificial neural networks (ANNs).

We often also use other terms to refer to ANNs. In the field of deep learning, the term artificial neural network (ANN) is used interchangeably with the following:

• net
• neural net
• model

What does deep mean in deep learning?

To understand the term deep in deep learning, we need to first understand how ANNs are structured. Once we know this, we will be able to see that deep learning uses a specific type of ANN that we call a deep net or deep artificial neural network.

In the next post on artificial neural networks, we'll learn how ANNs are structured, and this will give us the knowledge we need to understand how an ANN becomes a deep ANN.

For now, here is what you need to know:

1. ANNs are built using what we call neurons.
2. Neurons in an ANN are organized into what we call layers.
3. Layers within an ANN (all but the input and output layers) are called hidden layers.
4. If an ANN has more than one hidden layer, the ANN is said to be a deep ANN.

In summary, deep learning uses ANNs that have multiple hidden layers. Keep this in mind as we move forward in this series, and it will become more clear as we continue our journey through deep learning. See ya in the next one!