Understand Deep Learning in Artificial Intelligence

Learn how Deep Learning is used and how it works. Deep Learning is a type of Artificial Intelligence used for image and sound analysis and other types of pattern recognition.

You can binge-watch both  season one and season two right now.

Subscribe to the Lucid Thoughts channel and be sure to leave your questions and comments on each video.

Transcript:


Did you know the human brain weighs 3 lbs? And has over 100 million neurons that are constantly firing in it?

So recreating that on a computer should be pretty simple, right?

It’s not simple. But it is what we call deep learning. Math essentially simulates the neural network that we all have in our heads.

Neat, right?

Well, it’s more than just a brain model, at this level of training computers can identify and understand things in context. Like, if we give the computer hundreds of cat photos to look at, now I know, heaven right? It will learn to identify a cat.

But it runs deeper than recognizing pictures of cats.

You know how it’s pretty common these days to have smartphones and speakers you can talk to, and have you ever noticed how they get better at understanding the way we speak? Maybe you have an accent or you like to say words in a weird way.

[speaker two] Tomatoes.

[chuckle]

Well, as the computer learns a pattern, it also learns different variations of that pattern. So when it comes to voice recognition, that means it learns what you meant to say.

Deep learning can have amazing benefits in so many industries, but particularly, in the world of finance. This deep learning allows machines to fully understand you as a consumer. So as it monitors your transactions, it can quickly detect when something’s off.

This has changed the way we detect fraud and identify theft. There is a laundry list of pros to deep learning, but it does have some cons. For one, it’s really not that simple as looking at three examples of transactions, the computer needs tonnes upon tonnes of data, and even then you run the risk of overloading it and making its conclusions too narrow.

Remember that cat thing? Too many photos of cats, I know, I didn’t know that was possible either.

But it can actually give the machine a very narrow view of cats. This is called over-training, and when this happens it’s a lot harder to debug this model than some of the others we’ve covered.

But the beauty of deep learning, is the computer can learn new concepts on top of what it’s already learned. Like I said before, each method of artificial intelligence has its place in business today. It’s a matter of knowing when to use a fork or a spoon.

Now that you understand the tools in your kit, you can get started on building whatever you need for your business.


 

Related Videos
Virtual Assistants: How Question Answering Systems Help Siri, Alexa & Google Home
Question Answering Systems: How Search Engines Answer Questions Right Away
Query Intent Classification: How Does the Internet Know What I Mean?