Hi Guys.
Welcome back
Today we are going to see the basic definition
and then what is the actual difference
between Deep Learning, Machine Learning and then Artificial Intelligence?
These tech buzz words have been ruling the industry
and we are not understanding it properly
and we are actually using it in every places
because it is highlighted by most of the industries.
So now we can go to the video and can see what is the actual meaning.
Why we need to know the meaning of this particular three tech buzz words?
Because, first point is
these words and then technologies have now became integrated part of our life.
And second thing is
it has been the trending technology in every industries
and then every part of the companies which you are going to work in future.
Now let's jump into the explanation for this these three trending technologies.
First thing to be explained is Artificial Intelligence.
What is meant by Artificial Intelligence?
It is nothing but a computer system which mimics or replicates human intelligence.
Next thing what you are going to discuss is Machine Learning.
The simple explanation for this is
it is the ability that allows computers to learn on their own.
Next thing is Deep Learning.
If you see in Deep Learning,
it is the way of learning by formulating algorithms
in an attempt to model high level abstractions
in data to determine a high level meaning.
So now we will be going on with the examples of each and everything.
So that you can clearly understand its meaning
and how it can be applicable in a day to day life
and then in industry parts.
So let's start with Machine Learning (ML).
What will ML do?
ML actually processes the data.
After processing it, it actually analyzing and learning from it.
After analysis and the learning part is completed,
it actually applying these in real time scenarios based on the applications.
So what is actually happening?
The Machine Learning algorithm is actually being trained
or it's training from the set of data which are given as an input
and it is actually building its own logic to perform the operations.
So to do all this, Machine Learning actually builds the algorithm to work on it.
There are various algorithms which are being used in Machine Learning.
But they are being broadly classified into two categories.
One thing is Supervised Learning.
And then second thing is Unsupervised Learning.
Let's start with Supervised Learning.
What is meant by Supervised Learning?
As the name suggests, it has to be supervised.
I mean that the humans have to feed the set of data and then the solution for it.
But actually the machine will work on the relationship between two data.
So it will be very helpful in mathematical related processing.
So what is meant by Unsupervised learning?
As the name suggests, it does not need to be supervised at all.
And the machine will take care of all the functionalities.
We will be inputting the set of data and then the numbers.
And we will be asking to find the solution and the relationship between it.
So it is like, shooting in the dark and you don't know where it has been hit until you put the light on.
Let's start with the basic day to day example to understand the Machine Learning.
If you want to teach your computer how to cross the road?
The conventional method you will be doing it is
you will be writing the set of codes
How to take Left? How to take Right?
How to walk in the pedestrian?
and how to do all the things in a step by step manner.
With Machine Learning, instead of writing this code,
you can actually feed your system with
ten thousand videos of how to actually cross the road safely.
And the ten thousand videos of how you can actually hit by a car.
So based on this, the machine will actually train itself
whenever it is actually recognizing and working on how to cross the road.
From what we have understand till now related to the Machine Learning part,
it actually cut down the major part of writing the codes and then the logic in most of the functions.
So it is actually helping us to make the life easier
and then to make the implementation very much easier.
Having seen about the Machine Learning, let's start with Deep Learning.
The basic difference between the Machine Learning and then the Deep Learning is
the amount of data you actually feed into the system.
The Deep Learning actually requires large set of data to process it
and to give the clear and then the good output.
So if we have less number of data, you can actually go with Machine Learning.
And if you have a large set of data, you can actually go on with Deep Learning.
Another major part to be considered with Deep Learning is
it actually requires a high end processing hardware to function
and then to give the high amount of data and then the output to the system.
But be aware that, Machine Learning algorithm doesn't need high end processing unit.
But to process the high amount of data,
Deep Learning algorithm needs some high end of hardware and then the parts to work on it.
Why it is so?
And what is the actual reason for these high end devices to process in Deep Learning?
The first point is in Deep Learning, most of the part and calculation happened is Matrix Multiplication.
So that, it needs high end Graphical Processing Unit.
Second thing is it actually not breaking the process.
It actually giving the end to end solutions in whatever processes it's doing.
The third point is it actually provides the high value and then the detailed output.
Now, let's go with the best example for understanding of Deep Learning.
Let's assume you want to train your machine to understand
whether the object standing before it is cat or not.
For that, you have to work with different layers in Deep Learning.
The layer defines the neural network which is present behind the concept of Deep Learning.
What is meant by Neural Network?
Let's assume your object need to be identified whether it is a cat or not.
For that, you will be taking the cat picture at different angles
like its leg, claws, face at different versions.
And after that you have to place it in a layer by layer of environment.
Because in neural network, it has to cross various layers from the top to the bottom
before identifying it as a cat or not.
Then the networks also have to be trained with various pictures of different varieties of animal.
Because we are going to train the machine to understand whether it is a cat or not.
For that we also have to give the false input to the system and make it understand it.
After giving these inputs, the machine has to be trained.
For that, we will be giving the feedback whenever the result is provided.
The feedback can be given either by the human intervention
or it can be taken by the pictures which is available in the internet itself.
So based on the feedback, the system will be refined day by day based on the processors.
So at the final, you will be getting the clear output of whether it is a cat or not.
So that in Deep Learning, lot of processors and steps are involved to make it as a clear output
but the process and time taken to give the output is very much large
compared to the Machine Learning which we have previously seen.
Now let's move on to the next part which is nothing but Artificial Intelligence.
This part is something which is very much tricky to explain.
Because the intelligence level, what we considered as a mark
or appropriate limit has been changed over the period of time.
Before some years, we considered the systems which are playing chess
and it can do facial recognition have been considered as an Artificial Intelligence machine.
But now the mark has been changed.
The systems which can do the performance
other than which has been previously done
now considered as a limit for Artificial Intelligence.
So, now you can think that the Artificial Intelligence will never happen and it will not attain a limit.
But it is actually not.
The Artificial Intelligence have been defined based on the work it has been done.
Now let's assume the enormous cars have been given an option to work in a pooled network concept
which is nothing but if one of the car in the network learns a new thing,
it will update to the network
and every car which is connected in that network can learn it and then can upgrade it.
Now this is now has been defined as a Hidden Intelligence.
So that the intelligence level has been keep on being evolving one another the other
by the day-to-day life and day-to-day development is happening.
So this Artificial Intelligence have been changed over the period of time
and it has been evolved to the greater extents.
So these are all the three new concepts which we have seen today.
Let me know if we want to come up with a specific video of defining
various algorithms explained in Machine Learning or Deep Learning.
So that based on that we can deliver the content to you.
If you like this video, do give a thumbs up.
And if you have any comments, do post it in the comment box.
And if you have not subscribed to our channel, do subscribe it.
And click the bell button to get the update
whenever a new video has been delivered in our channel.
Thank you.
We will be seeing in the next video.



For more infomation >> 2/24/2018 It is fun to read the extra golden news on a PAPER than the tube/CRT. - Duration: 4:53. 






For more infomation >> Is it PHOTOSHOP? REALLY ? (Onedio) - Duration: 4:49. 

For more infomation >> Is It Possible To 'Catch Up' On Sleep On Weekends? - Duration: 2:13. 

For more infomation >> Denholm's team talk in IT crowd but the m in team is gone. It sounds like he is talking about tea - Duration: 0:18. 
Không có nhận xét nào:
Đăng nhận xét