Today in this Article, I am going to introduce to the idea of Artificle intelligence And Deep Learning and what it sort of means in simple terms right?

Now Artificle intelligence, let me break the myth That it not really a thing. It’s it’s a sort of marketing term that a lot of people use to garner attention.

But there’s not really something called Artificle intelligence as a science to study. So there’s, there’s Machine learning, there’s something called Deep Learning, there’s something called automation and rule based engineering.

So all of those are sort of subfields of data science. And that’s broadly the science that exists, there’s nothing really that you can actually call Artificle intelligence broadly.

Today, I’m going to try and, you know, sort of separate out what is Artificle intelligence, so called in the industry, what it sort of means, right?

What is Artificle intelligence?

What is Artificle intelligence in industry terms industry, actually, when it’s talking about Artificle intelligence, specifically, late 2016 and onwards, I mean, it’s actually referring to deep learning.

That’s what is normally referred as Artificle intelligence. And that’s sort of a hidden code language that we use as data scientists.

We don’t tell the world that that’s what is called Artificle intelligence. So when you normally hear today’s data, scientists talking about Artificle intelligence implemented in a system?

Probably this  what they’re almost always talking about is deep learning right?

What is deep learning?

Now, what the heck is deep learning? Right? So Machine learning is sort of the field, which extends into deep learning.

Now, to give you an idea about why deep learning came along. And why is it such a hype? Right?  so let me just give you an idea about what actually happened when Machine learning, right?

So Machine learning, for example, we take the case of Algorithm trading, right? People were trading in the floors in 1900s.

And then people someone said that, a let’s try and automate the entire process, if I can just give up some rules to the Algorithm say that, hey, you know, if the price exits this much particular price, or if the volume exceeds this much volume, just buy your stock, right?

So that was rule based trading. Now, Machine learning came along and said that, you know, what, you don’t need to set the rules yourself, just give me an entire feed of data.

And just give me which all trades to enter in which all trades not to enter. And then I’ll figure out, you know, which trades to enter and not to enter, right?

So I learn the rules myself, you don’t have to decide for yourself, just give me enough data. But then there are a small problem with Machine learning, which is that which all if it has to look at 15 different things to combine those rules, you have to specifically mention all of those 15 things as part of your data.

For example, if there’s something which is say price into volume, right, which is some feature that you want to look at while making a decision, you have to combine that as part of your Machine learning Algorithm.

Now deep learning sort of changes that rule, it now suddenly says that, you know what, you don’t even have to give me the entire sort of all all the data features that you want to look at for making a decision.

Just give me the basic raw features. And I’ll figure out how to combine them, and how to set a threshold on each one of them, and then make the most powerful Algorithm for making a trade right.

Now. That’s, that’s what is magic, right? Anything that is sort of, which can just look at the raw data and kind of give you the most powerful trading Algorithm.

And I’m just talking about trading for an example. Right? It is sort of similar example, for medical imaging, where watch, which is what I do right?

Earlier, you would have to write does this rule based system say that, you know, this pixel looks brighter, this is probably a tumor cell and all of that.

But the deep learning, you don’t need to do any of it, you just feed the image and say that there’s a possibility for that tumor, it’s possibly a cancer case or not, right?

So if you just do that, and the Machine just looks at a lot of data, it figures out which pixels to combine, and which way to combine and you know, and what is the threshold to set for each of them.

So that’s how deep learning sort of, you know, magically transformed this entire industry of salts, I mean, data science that has been existing since 80s, and all of that, right, Deep Learning came in.

And now, as you all know, is a region when you now normally here and probably now you would be able to appreciate that when people talk about why AI is so powerful.

That is probably because deep learning as a field is so powerful in itself. And when normally people are talking about, you know, ai chat bots, or AI based image diagnosis , face recognition.

they’re almost most of the time talking about deep learning based systems. And what is deep about these deep learning systems? Well, for that you need to sort of get technical but broadly,

if I have to explain you, it’s basically the idea of, you know, it’s a pattern Recognition engine. And you sort of combine the patterns in multiple ways. And that sort of becomes deeper.

If you combine, various patterns you can capture pattern in a very small level, keep on combine them, and, you know, building out bigger patterns, and that’s sort of what is deep learning.

But again, you’re welcome to join the course goes through the details. And if you have any further questions on what is Artificle intelligence,And deep learning, write in the comments below, we’ll be happy to answer them.