Artificial intelligence, deep learning, and machine learning are terms that are used interchangeably these days.
But what do these different terms actually mean?
Should they be used interchangeably or are they actually different?
Let me answer that question right off the bat, AI, machine learning and deep learning all refer to different things. The three are very closely linked, but they are not the same.
The below diagram demonstrates how artificial intelligence, machine learning, and deep learning fit together.
From the diagram above you can see that artificial intelligence is a catch-all term for many different areas.
Essentially, artificial intelligence is a term applied to a machine that is able to perform tasks in an ‘intelligent’ way.
There are two main types of artificial intelligence.
The two primary types of AI are general and applied.
In general artificial intelligence, the machine is able to perform multiple different tasks in a way that is deemed intelligent. This is like the intelligence that a human being has.
For applied artificial intelligence, the machine is able to use the artificial intelligence to complete one specific task or make predictions.
This applied artificial intelligence is the most common of the two types. Applied artificial intelligence is where machine learning, as you will be most likely to have encountered it, comes into play.
Before we move onto machine learning specifically, there is something else you should know about artificial intelligence.
Two further types of artificial intelligence run in parallel to applied and general artificial intelligence.
More types of artificial intelligence
The two types of artificial intelligence are weak and strong AI.
Weak AI is where the machine is intelligent in that it can take on information and react. However, it does not really ‘think’ or learn for itself.
An example of weak applied artificial intelligence is chatbots.
Chatbots are able to take in your data, most common questions, and provide you with an answer based upon their database of responses.
Machine learning, on the other hand, is an application of strong applied artificial intelligence.
What is machine learning?
In Machine learning you give a computer a data set, and it learns from that dataset to provide an output.
There are a variety of different machine learning algorithms that you can apply to the data set that allows the computer to learn about it.
The basic principle of machine learning is predictive analysis. You use the computer processing power to apply a predictive algorithm to your data. The algorithm will run and you will get a predicted output based upon the input variables.
From applying this algorithm, the computer can learn which features of the dataset input variables, are essential to the output variable, or prediction.
This output or prediction can take multiple forms.
For example, you can have a value prediction, a classification prediction or even a cluster/grouping prediction.
The machine learning algorithm can analyze vast amounts of data at a far higher speed than a human being would ever be able to. This is what makes them so powerful.
So how does deep learning come into play?
What is deep learning?
Deep learning is a specific type of machine learning algorithm that can be used for certain kinds of problems.
In my previous article, an interview with my colleague who is a machine learning engineer, we talk about how to use deep learning.
As you can see from the diagram at the start of this article, deep learning is a subcategory of Machine Learning. You use deep learning in the following three areas of machine learning:
In Deep Learning the Neural Network has multiple a lot of layers. This is where the term ‘deep’ comes from.
This deep structure of multiple layers allows the model to process more data and make predictions on more complex data sets.
Ok, so I hope this article has cleared up some of your confusion. You now know the differences between artificial intelligence, machine learning, and deep learning.
As you can see, all three of these technologies are very closely related but are different.