The Artificial Intelligence Evening

Thank you

Thank you

So very few students have been joined. Please tell everyone to join fast

Because we are expecting the pro vice chancellor also going to join with us

So I want that before that the students please join. Dolby, please connect with the first year students to join fast and the Rohit

Sir, OK, it's a no Robin and Sahil. Sahil have not joined till. OK

uh robin please circulate to the fifth semester and the third semester to join fast okay ma'am

Thank you

I have disabled the mic and the cameras for the attendees

Fine, sir. Fine. Fine. No problem. No problem. If attendees if you are having any query related to that, so please put it

into your WhatsApp group of the college. The presenters will speak from that. Okay? So

you don't get worried. We will be together for long. And you can ask your queries through

your presenters. Don't worry about that. That will be represented by them only

Thank you

Thank you

Thank you

Thank you

Thank you

Thank you

Yes ma'am. Hello, it's Sir. PBC Doctor Pradeep is also expected to join

I feel find the name with guys. You also please be careful when Doctor

Pradeep will be joining. Please make sure that we should all aware about it

OK, I make sure as a presenter also. Yeah, sure. guys please tell everyone to join fast as uh hi stephen good evening

uh hi dr palavi thanks thanks for uh the invitation uh how are you doing today

Yeah, we are doing great. Welcome, Simon. Thanks, Atul. Thanks. So you are going to motivate our future champs today

I'll just try. Yeah, I'm expecting. I'm expecting because really they wanted to join you

And we already got to know about your achievements from Mr. Gupta

So the students, I just wanted to tell you, Stephen, the students are from the, you know, first year, third year and second year

So they're actually really wanted to get into the AI. And when we told them that we are going to have an evening of AI, so their, you know, expectations are increasing

Okay. that what exactly because the first year students you're making me nervous

yeah we have to actually because the government is also working on hard on it

so that's why okay just give me a moment because i'm expecting more students to join

then we can start i know that you are very busy but i'm really sorry yes yeah more than 50 students

already joined ma'am it's good okay it's okay so can we start uh it's okay okay satak you can start

now over to you yes ma'am yes ma'am yes ma'am very good evening to one another present here

very good evening sir very good evening ma'am good evening satak good evening satak you just continue yeah so the title of the session today

is the future of AI. So this reminds me of a quote that is a year spent in artificial

intelligence is enough to make one believe in God. So here I start the session description

So innovation hinges on our ability to see the world differently. In this segment, join me to

learn how Microsoft approaches innovation, experience some of the projects pushing the

limits of science and engineering, and learn how these innovations will empower you to

transform your business or even the world around us. I would like to thanks to Dr. Pradeep Kumar

sir, Dean FIT MRIARS and Dr. Tappas sir that is HOD CAC department MRIARS and the coordinator for

the event Dr. Pallavi Goelma for this extensive webinar on AI. Now I would like to introduce you

all with our today's speaker Mr. Stephen Simon, the regional community director at C Sharp Corner

organizer of Cloud Summit that is world largest cloud event focused on Azure

two-times national finalist for Microsoft Imagine Cup, works very closely with different teams within Microsoft Seattle

Now I would like to, yeah. Mr. Stephen, sir. Over to you, Stephen

Yeah, I hope you can see me. Thanks so much for the invitation. That was a great introduction, Sarthak

And thanks, Dr. Pallavi. I see many people joining and I almost see there about 50 plus

So thank you so much, Aaron, for joining today. I am really, really excited. It's evening at my place

Also, I'm joining from very nearby Delhi capital, actually. And yeah, next 25 minutes is going to be really exciting

Just Dr. Pallavi told that the students from first year, second year and third year

So I'm going to try to share what I have. and maybe I can go ahead and start sharing my screen, right

Okay. All right. Perfect. Can you see my screen? Yeah, I believe you all can

Okay. So, hi, everyone. Once again, my name is Stephen Simon. I don't have any intro slides

You won't even find slides that have my name or you won't even find slides that have my designation

So, that's it. So, today we're going to talk about next 25 minutes about AI

Now, before I actually go ahead and talk about AI, machine learning, custom vision, computer vision, I want you all to go back

Go back to the good old days. Remember, remember, back, I'm talking about somewhere back in 2012 or 13, we used to go to this cyber cafes, right

We didn't have interns at our homes. We used to go to cyber cafes, use internet over there, and internet just won't work

The internet would be very slow, and we had to pay around 10 rupees or 15 rupees to use that internet

Now, that time, we didn't have smartphones at our home. Not even, there was no Wi-Fi

I remember I had a Facebook account, and that Facebook then and now looks very different

And today if you keep yourself updated Facebook has not only changed their name they have moved to a different ecosystem that is called Meta It more inclined towards metaverse I definitely encourage you to go ahead and see what Facebook has recently made the changes They moving more towards AR VR

and 3D characters online. So those are the days when we used to go to the internet

go to the cyber cafes, pay as less as 10 or 15 rupees, and use internet for 30 minutes

Now, why am I sharing this? I don't know. I'll come back to it, right

But I have one more slide for you. How many of you remember this

I don't think so. There's an option for chat. I don't think so. One can have a chat in this Teams call

But how many of you actually remember this? You know, this goes back to my childhood of nine

9 and 3 4th of the station. This is actually from a scene of Harry Potter

This movie, this entire movie series is very, very close to me. Whenever I get time, I always

go back and re-watch it again. I would highly encourage you, if some of you have not seen it

please go ahead and watch it. It is wizards and magic and all

but once again why in this world am I sharing this I'm coming to that let's move to the next slide

next month that is in just about a couple of days Harry Potter is going to complete 20 years

can you even imagine it 20 years of Harry Potter will be completed I still remember going and

buying CDs and DVDs to watch these movies. But now if you see, you have this Netflix

you have a great internet bandwidth. So definitely not just the apps or smartphones, the IT

infrastructure has evolved a lot. I couldn't believe when someone would have told me 10 years

back that you can watch a five hours or six hour video streaming on internet. Also, if you, I believe

Everyone here in India follows cricket. We recently had a match with India and Pakistan

Unfortunately, the results didn't go in favor for us. But never mind

If you would have seen that match was being broadcasted on Hotstar

and more than 1.2 crores people were watching. And my LinkedIn was filled up where everyone was saying

that Hotstar has done an amazing work, that even with 1.2 crore people watching the India and Pakistan match

the website won't hang. The website won't stop working. But whereas I believe whenever there was university results in my college

the website just would hang. But still, they're still right. So if you see, the infrastructure has evolved a lot

And the reason I've shared this small story with you is, are you feeling old

Did I just make you feel old? Did I just age you

Did you just feel that you have been on this earth for a really long time

Did you see a lot of the technology changes right from when you were in fifth and sixth

standard to where we stand now? And go ahead and ask yourself, apart from using these technologies, have you made yourself

very much updated on how to actually work with this technology? That is a question that I will leave up to you after the session ends, that technology is changing, but have you updated with this technology to keep yourself updated, not just using it, but also going ahead and, you know, building products out of it

So if you see. When I talk about AI and machine learning, I won't say I'm an expert

It's a very it's a very big ecosystem of AI and machine learning

But if you see, and broadly talking about this Azure AI, and for those who do not know what is Azure, Azure is Microsoft Cloud Computing Platform

Just to be, you have AWS for Amazon, you have Google Cloud by Google, then you have, I believe, Alibaba Cloud, and then even IBM Cloud

There are many cloud services out there. So Azure is by Microsoft

If you see, the AI can be broadly classified into five services

Now, I'm not saying into categories, I'm saying in five services, because Azure is a product, cloud is a product

And the reason I'm emphasizing more on the cloud is because everything is happening on cloud now

Even if you want to build an app, you want to build a website, you want to do AI, you want to do ytics, everything is happening on cloud

So it would be great that if you start your journey in any ecosystem of the development and think of it from the cloud at the very first place

Now, the very first thing. Stephen, one second. One second. I'm just disturbing you

Actually, I want you to welcome our Pro Vice Chancellor, Dr. Padip Kumar, sir, also joined to us

Hello. Good evening, sir. Good evening, Dr. Parvi and welcome to Dr. Stephen

Sorry I interrupted. Please continue. No, not a problem. Thanks for joining in

I really appreciate it. I know it's evening time at your place. So, yeah, let's get back

And yeah, so what we're talking about is the Azure AI ecosystem. Broadly, it can be classified into five categories

The five services, actually. The first one is machine learning. The second one is anomaly detection

The third one is computer vision. The fourth one is NLP. And the fifth one is conversational AI

Now, this may look, okay, there are only five services to learn. But here's what I'm going to do now

Right? I hope you guys can still see my screen. I'll Google Azure AI

Okay. And I click on the very first link. Or maybe not

Not the first one. The first one was ads. But never mind

Let me just Google Azure. Okay. And then I go to the website

Okay. Under products. You see there are many products out there. I click on view all 200 products

And the very first category that you see is the AI and machine learning

Now look at the different services that you have. I don't even know how many are there, but there are a lot of services and things are

evolving very, very rapidly. And it is very challenging for me to go ahead and cover all this

So that is why what I've done is I've taken those five important services that one should understand

The first one is definitely machine learning. Now, machine learning is definitely the fundamental of AI, right

If you want to move into machine learning, the two most famous programming language is Python and R

If you're someone who's just getting started with machine learning, the very first thing is to actually go ahead and learn the programming language

The second one is to go ahead and learn some of the frameworks. Now, some of the framework means either you want to learn scikit-learn or you want to learn TensorFlow

I would probably say scikit-learn to get started, then probably move into TensorFlow

And once you know on how to actually go and work with the algorithms one step that many people miss and I want to go ahead and really emphasize is learn on how to do all the algorithm from very very scratch That is important to to understand how this algorithm works Then you whatever then go ahead and build a small project

and then put it on cloud. That's a bird eye view on how you would build a machine learning project

So machine learning projects, now here's one thing, right? Whenever I say Python and R, there's always a confusion

between whether I should choose for Python or an R. Here's what the industry would see

That if you come from a developer's background, if you have been coding and all

then preferably you would prefer to use Python programming language, which is absolutely great

Whereas if you come from a database background, if you come from the ytics background

you're a business yst, you are a data yst, you work with charts and all

then you would incline more towards that. And both are great. I'm going to talk about ytics if I get time towards the end

because that domain is in demand a lot these days. ytics are the one who act as a bridge

between developers and the data scientist. So yeah, that was the machine learning

a very, very gentle idea. The next one is normally direction. Now, here's what something, for instance

you work with a data set. Now, data set is actually the

you would say the basic input you would give, but the very first step you start working

with the machine learning and your job is actually to identify some of the things which

are not normal. For instance, let us take this scenario. A student is performing very

well. Then all of a sudden in particular semester, he starts failing. So with anomaly detection

I can actually go ahead and figure out that, hey, this is something which is abnormal with

the students. Now, imagine I just said it about a school. What government of India and

many people around the world they do is they look this for the suicidal cases

All of a sudden, student is working very fine, then boom, students commit suicide

So these are the things that government have always started using it to implement, to take

the best out of it and making sure they use the normative

It looks a very small topic, but it is of immense importance

The second one is the computer vision. Now, here's what. I'll go back again

Remember, if you used Facebook, and even if I believe you all use Facebook now, which is now meta

What happened sometime back that if a picture was uploaded back in 2013 and 14

Facebook would start telling me that, hey, are you tagged in this photo

Whereas I was never tagged. That's what Facebook would do. So what is happening now

That is actually your computer vision. You actually train your machine learning model to go ahead and understand the images and to give you a little idea

Images and all, at the end of the day, they are zeros and one and they're filled with mattresses

When you move into the field of machine learning, you'll work a lot with mattresses and NumPy and SciPy and all that stuff

I'm not going there, right? I'm not going there because I just have about 12 or 13 minutes more left

So computer version is something that you should definitely go ahead and explore

And these are the five domains that I've added just now. You can actually go ahead and build carriers in each one of them

You don't have to learn all of them. You have to learn just one of them, build a niche and then start your carrier

The next one is natural language processing. Now, if you go ahead and have this Google Assistant or Alexa, the moment you say you can your Alexa or Google can understand

Now I can put an LP and conversation AI just into one. Okay. They can understand it. They can respond to you. There are bots on the website. So we humans can understand every single word that we say, right? That's how our brain has been trained, but not the machine learning, right? So we have to actually go ahead and train it from the very scratch. And I think every time people would join, there's always a pop notification that keeps joining up. Okay. Never mind

Okay, so that was an overview of AI, okay? But again, as I said, there is a lot that needs to be covered

If you see, if you go to this AI and machine learning, there's so much of things

The anomaly detectors, the Azure port services, the Azure open data sets, face API, Q&A makers, speech to text

So definitely go ahead and check this. I'm going to, I cannot drop it in the chats

But if you go to Azure, as I said, if you go to Azure, OK, and click the products, click on view all products, then from the left AI and machine

And if you just click on it, you'll find all these different services. Now, these are very easy to use. This is what the enterprise would use

It you're a student, you're a first year student. So if you want to go ahead and learn Python, it's OK, it's totally fine

But as you progress, definitely start visiting these cloud platforms that will give you an idea on what are the tools that the enterprise is using

OK, the sooner you start, that's better. The other thing, and I want you all to stay towards the end, because apart from this, in the last five minutes, I'm going to share some of the very important resources that has helped me to grow my career

So just definitely stick towards the end. The next one that we have is the low code and no code

Gone are the days when developers would go ahead and build applications from very, very scratch

What you see right now on your screen is a machine learning model

We are adding the data. We are doing feature engineering. That is, we are removing all the bad data

We are removing the column that doesn't have some values. We are removing the repeated data that we have

And all that is happening without writing even a single line of code

Now that is where the world is moving is to how we can go ahead and remove all that repetitive task a developer would do in a day to day day to day work life

So it is very very important that you also keep an eye on how you can go ahead and build low code and no code applications

Now what you see on the screen right now is is is a GIF or GIF

What do you want to say from the Azure Machine Learning Studio or designer Azure Machine Learning is once again one of the products from the Azure AI

ecosystem that allows you to go ahead and build end-to-end AI and machine learning applications

You can go ahead and build your machine learning models, train them, use the compute that is on

the cloud, get some endpoint that is APIs, and then actually connect with any of your websites

or the mobile apps that you have. It may be very overwhelming with all the information and the

technological words that I'm using for the first-year students, but I want you to go ahead

to and actually watch videos on youtube take some courses online to get an idea on how uh

the things are working out okay so definitely go ahead and check out the azure machine learning

studio and the other one that i have is the power path powerful alone okay that that's a

that's not the heading the other thing that i want to talk about is the power platform now you see that what you saw over here was actually a low code no code for machine learning

But, you know, the entire developer ecosystem has been changed drastically. If you want to go ahead and build a mobile apps, you need not write the code

If you want to go ahead and build a website, no need to write a code. If you want to go ahead and build an admin dashboard, you need not write a code

If you want to go ahead and do anything, you want to go ahead and build a bot, not a problem

Don't write a code. If you want to automate a process, no problem. Don't write a code

So you see, it is time when you go ahead and learn all these intelligence tools that's there in the market

It is very, very important that you keep yourself updated. For instance, right now what you see on a screen is Power BI

Now with Power BI, you can actually go ahead and build the dashboards

These dashboards are used very much in enterprise. You need not know any programming language

It's all drag and drop. In the same way, there are, if you want to build mobile applications, you have to use power apps

Then if you want to automate things, they are power automate. And if you want to build bots, they are power virtual agents

And each of these services, each of these products have AI capabilities

If you want to train a machine learning model, you don't have to write a Python code. You don't have to write an R code

You just have to upload an images. These services will train on its own and you'll get a machine learning trade model

That is where things are moving. So you have to be very, very smart on what are the tools and technologies that are happening in the market

You may end up learning TensorFlow from very, very scratch. You may end up become a very great developer

But once you graduate, you see that there are very less jobs in TensorFlow, whereas there are a lot more jobs in all these new tools and services that's there in the market

OK, so that that is important. Now, before I go ahead and move towards the second part, maybe what I can do is let me quickly change my screen

Okay. Okay. Maybe I'll move to the next one. Okay. The next thing that if you have reached this part of the event, I still see many people are joining in

Right. So definitely they have missed the first part of what I have covered

So here's what if you have missed whatever I've said today until now and it's 23 minutes, I have about five to seven minutes

If you have missed, it's OK. I want you to forget whatever I said. I really don't mind

I really don't mind. But the next three to four slides that I have for you, I want your full attention

I literally want your full attention on next three to four slides that I have

Forget if you want to forget everything that I said, forget it. Keep your phone the site

switch off your notifications. Next five minutes, I'm going to help you on how you

can shape your student carrier. The size that you see boys and

girls on your screen is a program that's called Microsoft Learn Student Ambassador

Back in 2014 when I was in college, I was a part of this program and to

what I have achieved now in my career, I give a lot of credits to this program

If you become a part of this program, you officially represent Microsoft

in a college campus. How cool is that? Not just you represent, you actually get the free services

like you get Azure, you get Windows and a lot more stuff. You get to go ahead and build connections

with people working at Microsoft like Cloud Advocates, Senior Program Managers. At the same

time, you also get an opportunity to attend these free Microsoft events. Now it's virtual, but then

things were in person. And I still remember back in 2014 when Satya Nadella, the CEO of Microsoft

visited India for the very, very first time after becoming CEO. It was 30th September 2014

I was there. I was invited to meet him. Not just one, I met him thrice

So it is all about this program. Microsoft Learn Student Ambassador. I'm not sure if the applications are open at this stage

Most probably they should be. It is fall, so they accept the application

So go ahead and apply. It's not very tough, but this will definitely help you shape your career

The next one I've been talking about Azure, right? But Azure is paid

To be honest, Azure is not free. It is paid. But you can actually go ahead and get a free Azure subscription

You get free popular services for around 12 months and you get plus another 25 services

that would be very, very beneficial for about 30 days. Plus you get $200 USD credit

That's where you get it. So everything that you go ahead, what you saw, that custom vision, computer vision, machine learning studio, the bot services, you can

actually go ahead and practice that all using the free service. So just Google Azure free

subscription, it will take you to the link and it is really going to help you out

The next one is the Microsoft Learn. Now I know my sessions are inclined more towards Microsoft

technology, but that's my background. So if you want to go ahead and if you're

missed everything that I have said right in this session, if you have missed, even you have missed

the slides that I've shared right now. So Microsoft Learn is a platform you definitely want to visit

every single day. It's a platform where you'll find free courses, right? You'll get to know about

certifications. You'll get to know about Microsoft latest live shows. You'll get to know about

free events that are happening. So Microsoft Learn is a one-stop platform for every student

and developers out there to do all these free courses you can see you can get to know about

certifications you get to watch all these live and recorded sessions and many times you will find me

on this platform doing live shows and conferences i i come almost like every week at the i still

remember september was like the entire month i was on this website almost like 12 hours each day

i was streaming okay the next thing that i have is i want you all to go ahead and look for

certification even if you're in first year believe me i did three certifications when i wasn't in

first year things were very different than i'm telling you right we didn't have all this internet

connection i would get very happy when the students would bunk the lectures right and then i would go

to the lab and use internet the system would take 10 minutes to switch on and then i had to switch

off the another 10 minutes and i would get only like 15 20 minutes to use the computer so you guys

and girls are very lucky that you have internet connection so here's what the uh if you if the

entire azure certifications looks like but this this might be very very overwhelming so what you

can actually do is i'll quickly i'll quickly uh search microsoft certifications and uh

click on the first link and if you want to become this is an ai session right i'm going to click on

you see how to become an AI engineer. I'm going to click on it and then it is going to tell me what

are the certifications that I need. The first one is Azure fundamental certification, then the other

one is Azure AI engineer associate. The first one is not mandatory, so you need not take it

And let's say, suppose I click on Azure AI engineer, it is going to tell me what are the

skills that is going to be tested and over here, right over here, you find all the learning tracks

that you need to learn to pass the certification. How beautiful is that? And this entire course is

is free. Okay. This entire course is free. Okay. I just have about two minutes. There's also one

very important thing that I want you all to know. Let me go to the certifications, browse the

fundamental certifications. There's a very important certification that I want you all to

definitely do That is Microsoft certified Azure AI fundamentals Okay This is a certification that I want everyone you all to go ahead and do it Okay it very very easy very very easy very very fundamental

And you see, I'm going to show you the price that your base is $3696 plus taxes

Now, that is a sad part, right? You don't want to pay it. I mean, this $3696 still looks me expensive

I don't know, you are a student. So I'm going to give you a trick. You can get this certification for free

Everyone in this call, you all can get this certification for free. And how does it happen is you have to go to Microsoft Virtual Training

Okay. I just Google it. Use any of your favorite search engine. Okay

One minute. Microsoft Virtual Training Days. You click on this. The very first thing that you get

Trigger, trigger, trigger. Oh, my goodness. What happened? Hold on. Hold on

The last thing you want is Microsoft website. Now, yeah. Oh, there you go. So, here are the virtual events

So you see, you have to find all the trainings that they're giving and all these trainings are for free

And if I just scroll down, you see there's one virtual training that's happening that is on AI fundamentals

That's happening on 11, 18, 2021, 11 a.m. Pacific time, which is about, I would say, 1130 p.m. Indian time

But you can actually go ahead and sort it according your time zone. OK, there are many you can actually find many other times

So what you need to do, you need to click, you find this AI fundamental training

You click on this register. Okay. This registration is entirely free. And when you complete this training, which is just for about, I would say how long it says

It is about, let me check, let me check, 11 to 2

It is just about three hours. Okay. This training is just for about three hours. And once you complete this training, it's just that video they teach you

Once you complete the training, they will give you free exam vouchers. And here's a secret thing, right

You don't have to complete that training. you'll get the voucher just the moment you log in that's the secret okay so you can use the

microsoft learn platform that i've just showed over here don't tell them the shortcuts please

yeah so if you actually you don't have to actually spend three hours just learning it right if you

actually go ahead and complete all these resources that you see over here you're good enough to go

ahead and complete that certifications okay so i believe that's it i only have about oh i've

exceeded one minute so i think that's it everyone thank you so much for joining in it was really

amazing having you uh during 25 minutes i hope you liked it and yeah thank you so much and hope

to see you soon thank you so much stephen thank you so much to make to assure to my students that

how they at least can get the free trainings also because in the 50 minutes of the lecture we

sometimes skip certain part and thanks a lot that you guided but the shortcut which you shared that

i don't like being a teacher uh yeah thank you stephen hope we will get another series with you

soon yeah thank you so much and thanks everyone for the night thanks everyone for joining have a

great evening so it was a great session by stephen so i would like to welcome our next speaker

mr shubham so he's a date he's a data scientist he's an international speaker he's an author he

has written a very wide range of articles on c-sharp corner you guys can visit c-sharp corner

and I will look at his article. Over to you, Shubham. Thank you

Thank you so much for this opportunity to share my knowledge and also to leverage the kingdom of learning

to the students. And I really appreciate the management of the university. I really do appreciate

So in this session, what I'm looking forward is not just about learning

or not just about one-way learning, right? I'm just looking forward for an engagement of the students

In the middle, what I'm going to do, I'm going to ask some questions

It will be very beneficial or very good for me if an engagement can be done by the students

And the second thing that I can, obviously, I'm going to do because learning coding is more towards into practical of coding

So I'm going to show you not just about what is AI or what, et cetera, et cetera, but I'm going to implement the same on practical coding itself in one of the applications in the real time right now

So I hope that the sessions can be benefited for the students

And also, I would really appreciate if you do have any sorts of doubts

feel free to raise your hands and I would be gladly to over to that person and I'll be answering the very same questions at the same moment

So allow me to share my screen for a moment, please. Please give me a nod if my screen is visible to you all

yeah your screen is visible to us yeah thank you so much thank you so much i think that you have

got the definition of what is artificial intelligence it doesn't need to be in a device

and simon have delivered you a wonderful thing and conveying what ai is and what ai can do and make

things but for me because the source here i have mentioned it's how artificial intelligence is

empowering technology right it's not just about learning ai it's more towards how it is empowering

technology which in fact you possibly are using it in your palm so to be precise the artificial

intelligence may be defined as an ability to think make decisions and decide the best options available

now i would love to introduce you the three thing here the main keywords ability to think

make decisions and decide the best options available that means the options would be

given an opportunity for ai to learn and upon applying the options it can make decisions

during the process of making the decision it goes through a process of thinking

now you can connect the words as an ability to think make decisions and decide the best options

available it adds to the function ability to identify to recognize to express emotions and

also convey messages and informations like human being in short ai is a machine's ability to add

human function ability thus complementing the best possible outcome of a situation and that is so much

you can convey to the real world scenario even today when you are using things how in the next

few slides i'd be coming to for now if you want to learn about artificial intelligence you can

hover to the artificial intelligence c sharp corner and you might you know uh love to have a glance

over my articles that would really appreciate if you do have any sorts of doubts like on those

articles feel free to comment in the section and i would follow up with the possible answers at the

next moment of time possible now the thing is we have learned about or we have gotten so much of the opportunities that have been made by AI what AI is what AI thinks Now it really important time for to understand or to implement

or to understand what is AI in actual way or if not how AI is building itself and what could be

the better tomorrow that I think. So AI is integrated into almost each and every aspect

of end user software experience if you are i think i think that your first year engineering students

or any other but you already have used softwares be it in your web be it you know web apps be it

you know androids ios etc and etc now the thing that you love about your software be it any software

be it from online using buying things or online food ordering you love the experience of how you

use the software first if you're for an example if you're searching for a food to buy an online

you need search the name of the food or the name of restaurant and then you click over to the

quantity pay your price and delivery and that's that's the beauty of the software experience right

and thus it helps an AI to build something let's just see how so you might have used

online shopping or online food ordering again to be saying at the end of when you're buying

you might have a recommendation from the store itself that if you love to buy something

If you go for to buy shoes, it may get you an example or a recommendation system of socks or something with that particular if it's a sports shoe, sports t-shirt, etc

Now, how they are doing things, it is the recommendation system. Recommendation system is very short in word

But trust me, the work behind recommendation system, it's actually magnificent. The question is, how do the people know what I'm trying to buy or what I love to buy

The first thing is gathering data from the past customers. Data collection

Using data ytics techniques to understand which things people are buying the most

Getting the best decision out of those data is what data science invests

and building a recommendation system out of those all the decisions includes machine learning

and when we cover all the things from data collection to data ytics from data science

to the machine learning to be very honest you're doing 80 percent or 75 percent of the work of the

ai so most of the thing you see online stores which tends to suggest some products enhances

about user experience with the side of the online store. Sometimes when you browse your online content

be it any streaming service, if you have ever used any streaming service

et cetera, Netflix or Hotstar or Amazon Prime, et cetera, et cetera, with respect

you might notice that there, after watching certain movie types or categories

it suggests you some movies or shows which might be in your flavor, right

So, yeah, these things are the user experience at its best, and services rely on a recommended system that enhances the ability

Second is best route service. You might be, you know, like used of the things such as Google Maps, or if you're going to a new place and don't know the route

So you search on a Google Map, and it shows you the path. Obviously, from point A to point B to reach, there are several other ways or routes, but

suggests you the best route. How? It is actually the real-time thing that data ysis and data making decisions gives on

the much of an integral part of a subset of AI. Third, and the very beneficial thing, is security service

Finding or recognizing a face is another subset of AI. with additional of an image recognition system

Here, I would love to convey a keyword here, image recognition system and other cognitive services

AI has so much of the applications. It is one of the image recognition system over here

Some services use a collaboration of hardware and software. How? Let's just see an example that if you use an image recognition system

for a security service. Until and unless you do have a camera

it's not going to work. And same wise, if you don't have the software

with camera you can't do nothing. And that's the beauty of hardware and software together

So, with algorithms and supported services, works real-time scenarios. One of the best possible

you know, like, possibly a radio tool. In day-to-day activities, you must be using your email, right

Email categorization. Now, what is that? For the first time, when I heard about a few months earlier

to be honest, a few years earlier, I was just like, what is email categorization

I'm not getting anything out of that name. Best answer is when you log into your account

you must be using the inbox. There is a spam. There is a junk, right

Folders. So how the software differentiates among those? Well, the contribution of a core AI is trained into their modularity engagement

and that contributes to them to understand approximately, not just 100% of the time

but approximately to understand what is a spam or junk and what could be beneficial for you as a promotion or a primary inbox of your system

and the very very beneficial recently it's been done by various banks

various banks recently it's it's just like about you know like one and a half years in the practical

thing if somebody wants a loan or credit card the bank actually using the ai to understand not just

about credit score but how much to give the person the evaluation of the loans and that is

natural deficience, saves time, saves human energy, and also increases the efficiency of the system

And as an AI practitioner, you have to understand that whatever today you are using

whatever data you are giving to the companies or organization or being collected, it is actually making the system more

and more proper in the way to be defined as artificial intelligence

If you ask somebody what is artificial intelligence, most of the people will just say, you know

like it's a way that makes a machine think like human. But what is that thing that makes a machine think like a human

It's data. It's totally the data. Put it in the algorithms training testing and then comes the outcome And that the beauty of a data And when there is a data huge data and there the followed beauty of a big data

So now, it is one of the sins I told you that I'll be following a practical thing

of a face recognition system. So you must be known about face recognition system

It's a system or a software that understands the face and recognizes it, right

So till here, I'm going to all of a theory I've done

So if you do have any sorts of doubt, I'm going to stop my sharing over here

If you do have any sorts of doubts, feel free to ask me because for the next moment from starting

I'm going to show you to build a recommendation system in a practical way

in your machine. So yes, please, if you do have any sorts of doubts

Uh, yeah. Hi, uh. Hi, Shubham Doctor Pallavi here

Hello ma'am. Hi, good evening. Thanks a lot. Thanks a lot. Uh, I think students are having certain questions

So you were having one question or Suryanch, please ask because students are studying AI and we are trying to tell them how you can become a good AI engineer

So here you can have the students from the first year, second year and the third year who all are having this subject AI

Because we are making to indulge them into the algorithms more because as their semesters are increasing

So the first year just at the introductory part of the AI

And as Stephen explained about them, that how you become an AI, how you can go for it

it's pretty interesting. Suryansh, you are the presenter, but you can unmute yourself and you can ask the question

Sorry, Shivam, the students are having certain questions. I'm Suryansh

Yes, Suryansh, you can ask the question, please. Good evening, sir. Good evening, Suryansh

Good evening, ma'am. Good evening, please. Sir, actually, pardon. Due to some error, I can't able to unmute myself

No worries. So actually, my question is how we can use AI to protect endangered species

Right. So there are several projects based on the same various category

What you can do is actually gathering the data set from the open source of certain forests or certain area where government is providing the data

If not, you can just ask or email the team about the same that you need the datasets

And upon the datasets, you have to understand, go through the manipulation of data ytics

followed by data science. Now, allow me to explain this thing. Why data ytics

Because when you are using the datasets, it gives you the calculations of what was the

number of that particular animal in the past year and what is this in the second year or

in the present year. Now, this data is not sufficient to give you the answer

What you have to understand is you have to collect the data from at least 10 years back around the area

And in that process, you're going to use the data science algorithms to understand the graph, to understand the trend of that particular species

And from that graph or the species based upon your prediction, you can understand and data science will give you the answer that the species or the number of animals are going downwards

Downwards means, you know, the number of counts itself or it's increasing

Using certain machine learning algorithms. What you can do after followed by data science, you can use certain algorithms, use the data science in training set, use certain methods

how the past things have been saved and thus you can contribute to how if certain species

are left or slowing down or getting endangered what is the most possible way to save them

that is one of a kind you know recommendation system that you're building for your own self

it's not just about finding the which pieces are the list that questions in the data science and

only but you are building a system in how they can itself be saved from the particular region or area

And those the compliments of an AI. Thank you so much. Appreciate it

Any other questions, please? So what kind of threats

why we can use this AI applications, what kind of threats we can face

I mean, for the risk management, if we really wanted to work on. Because we are completely on the cyber

So we might be facing certain issues because the students of the first year really wanted to know

They're excited to use the AI applications, but wanted to know what kind of threats they can face

Correct. It's everything in this world that I've been using, be it a mobile phone or be it a cell phone, be it any wire, be it any bat, be it anything, etc

and et cetera, has both a positive and a negative impact. And thus comes to the same as AI

AI is very much, to be honest, it's a very great tool, but it does have a negative things also

which I need to cover here. The threats that you will be covering in which AI

is the very first and the foremost things in which you might be hitting a back or a bad back

to be honest, is gathering the data sets. because all over your recommendation system

or modularity or even your algorithms are going to work on what? Your datas

And if the data is wrong, your whole plan from very starting to the end

it's not just a waste of time, but a waste of your interest time also

So the very first thing, if your journey is towards is very careful, very, that's the reason why most of the data scientists

or AI engineer practitioners or AI engineers spends their at least 30% of the time

in cleaning the data, if the data is correct or not, if the data is beneficial or the data is even relevant

for an example someone just asked me a few you know moments ago endangered species topic it was

if i want to find the number of elephants are getting endangered in certain region

or any other animal what if i don't get the data set for the animal of elephant what if i get the

data set of a tiger so the your whole planning goes into break thing so first thing first thread

you have to get the data set or clean the data set if needed or even filter it second thing you

have to get at least understanding of uh data algorithms be it anything machine learning data

science data and yzing or anything that you're building right so until and less you don't have

that understanding at least a theory understanding of why you should do that algorithms and why should

not. For an example, if you want to predict a student's marks, you're going for linear regression

And there is another certain type of algorithm which actually makes things faster than linear

regression. The question over here is, both are correct. Why should I search for them? Well, there

might be certain algorithms which computes fast for large data and certain less. And that's the

the beauty of the thing of every algorithm you have to understand the requirements second

efficiency third working time and i think that would be great so second second is algorithm

understanding and the third is the objective very important thing your objective without objective

of your whatever things you want to make out of the system you can't decide your

algorithms you can't decide your machine learning patterns you can't decide whatever things you want

implement in your code. Third thing is a threat I think I have personally faced

myself when I was just starting into this journey. Objective. You have to

understand what thing you want to bring from the datasets and that's the process

you will be following throughout the rest of the process. With that understanding, I think that is that is that answers the questions. Yeah thank you

So, any more questions please? Nice, thank you so much. I would like to add to Shubham

So as Shubham said about objectives, so I would take you to one example, like I don't

know how many of you know about the example that China created an AI that could vary the

genome or you can say the DNA so that DNA was banned so that machine was banned so their objective

was I don't know but was their objective but it was very harmful for the humanity because they

say that they can produce hundreds of Einstein with that system so if your objective is right

AI will never be a disaster. AI is a machine. I always say AI can never have what humans have that is humanity

We understand emotion. So if a person is fearing that AI will take over my job or AI will be harmful

it will not take until and unless it develops something, I would say like Chitty in robot

He had emotions. So that would be disastrous. See, Rohit, I just want you to interrupt

If you can see the all first year students are here. Dolby, I have used the same example in my class also

In during the AI lecture you know sir whenever we are teaching no Shubham the same things you were telling Yes we are giving the live examples to them because AI is exactly the same thing It is trying to read the human brain And it is behaving like that only But I always told to my students not to believe on all these things facts

Because yes, absolutely. You're on it. So this is the, you know, guys, we are actually, AI is trying to control it

I have given the example of LG or that refrigerator also, which is telling you that, which is the rotten vegetable over there

and we are fully going to depend on these machines. So same thing

I would love to see your notes also. Shubham, that Mr. Rohit already told us

that you are a good author. We will cite your research work also

So guys, any questions? I hope. Thank you so much. Should I just start with the practical now

Yes, we can. we can start perfect perfect guys be careful because you have a ai lab also it is going to help you

uh please give me a nod if my screen is visible to you

Yes, sir. Right, right. Thank you. Thank you. So just told you about a few moments earlier that I'll be showing you one of

the applications of AI that is image recognition system or a face recognition system to be very honest and very direct

Let's just see what I'm going to do here. So in this image, you can see there are people here

So they might be using certain, you know, like office work or etc. etc. or enjoying time together

So in this picture, what I'm going to do over here, I'm going to recognize their faces

I have to make my system understand what are faces and what are not hands and what are not boots, etc

At the end of this session or at the end of this practical session, what I'm trying to make you understand is

I'm going to recognize the faces, only the faces. And that could be used as an attendance system

or that could be used as security things to learn faces of different people

or to find something or build some database out of those things, right

So yeah, let's just see. So if you're a beginner in Python

or if you're a beginner in AI or et cetera and et cetera, so let me just teach you or make you understand

from very basics and that will help you out. So the first thing I'm going to talk about

this compiler, what is this? This is Visual Studio code and what I'm using here it is

a Jupyter Notebook extension. You can get various extensions on Visual Studio, right

Just clicking over here that would be great. Any languages you want it got here. Anyway

the first thing what I did I'm going to train my model into certain I'm going to test it

also in a certain you know like picture whatever picture i need and that's the reason i use this

picture i inserted the folder and the second thing that i'm going to do over here is importing certain

libraries here i am using the cv2 how i'm importing here just write import cv2 first thing

it is important for you to understand that use whatever you're using you have to understand

that you have to install the libraries first before doing how you're going to do

with if you have any sorts of you know like in the machine python the simple thing that

you going to use is just you have to write pip install and followed by you know like whatever you think so for this thing i will use the opencv so in this this way you can install it since i have already installed my machine that the reason i not installing

over here you have to open a command prompt with admin permissions and you have to run the command

and that's it making sure you first have python installed so i have inputted the library over here

the second thing i have to load the image so i have named a variable as load image.dpg because

i have renamed my image as image.jpg over here and the third thing that i'm going to do is passing

this image into the library or running into the library itself so it is here image here it is a

variable cv2 is actually this library dot i'm ready to the function where it enables machine

to learn or to read the picture or to accept the picture from the picture we wanted so passing the

load over here so this load over here i think you have understood so what i did imported the pic

imported the library second thing i have done over here is loading the image third is putting

the image into the library itself so here it is cv2 dot im read load third thing that you have to

understand is i'm going to you know tron you know like training the model in such a way

that it makes a classifier face classifier how did i do that so i'm going to over here

spatial this is the keyword that i used over here again calling that library cv2 dot casket

classify it classifies certain things based on the parameters delivered later in inside the brackets

inside these brackets so casket classifier what I'm asking the class get

classifier to again include the library asking the data to classify their

frontal face so it's an in building that you have to just call it itself so here

it is the very important thing is here you have because each and everything has

a different color you have to change into a grayscale so it becomes easy for

machine to understand with respect to any color because you're converting into gray scale to

categorize things excuse me so this is a gray scale over here and this cv2 dot color it actually

converts into the gray scale and third is I'm going to detect the faces not they are very well

here it is faces facial dot detect multi-scale that is i'm using a command over here that detect

faces how it is it is facial dot detect multi-scale i'm passing this gray scale

over here and asking them to build a rectangle itself next is building a face rectangle uh it is

very like easy or it is very you know xyz take it a variable in phases you just ask the you know

like a library build a rectangle from where the image x and y and this is going to give you a

rectangle space like it is building a face rectangle and the last thing that i'm that i'm doing and

that's the beauty of the thing it's so much less of a code you build a model and i'm going to show

you the output also of this model also it is you're asking the machine that to show the image

and asking the machine also to wait until i see the image excuse me so i'm asking over here cv.im show what image

and passing that image and cv2 dot wait key is asking the machine to wait for certain time

So let me just get a compilation over here Can everybody see the output

Please do. Let me know if the output is visible to you. So here it is, after moderating, issuing everything, here what it enables that there are so many

of the people, one, two, three, four, five, five of the people, it actually understanding

these faces. That this is the face that is a different part of the body

And that's how the machine is using its ability to understand faces. One of the most proficient way to understand what is a face recognition system

Under face, you can build certain types such as security things, certain attendance systems

systems or etc. One of the best beneficial things of learning Python is building your model into

crisp, short and on spot. So here it is. I think you guys have learned about it. I'd be happy if

you still do have any sorts of questions. I think I've done it. So yep, I'd be I'm open for questions

please yep yes please

I think Shubham they must be trying

on hands on then they will be able to ask something they were not

ready for that because it's better when we will be interacting in the next series

so they will be uh you know download it and or the online compiler they can work on it

right now they are just interesting to know and they're happy it is it is i really i do really

appreciate the university and also the management of the teachers for enabling that fire into the

students to be that future workforce i really do appreciate and i really appreciate the students

also because uh understanding your niche understanding your niche information is really what makes you stand apart from different students

Do appreciate everything, everyone. And I think that, yes, I'm looking forward for some questions. If you do have, love it

If still, no worries at all. I already asked them, I think

because they came at five to their homes. No, after that, they are on it

Right, right. No worries. No worries. Definitely, we'll be in touch with you

Soon, we will be having again one more series with you. As it's already on

Sure, sure, sure. So I hope for the today. Yeah, Rohitji, over to you

Thank you all. And it was a great session. Like there was people who are still sitting for one hour after their call days

So great, great. I really appreciate the students and the management. but especially for making this happen

Thank you, Shubham for being there. Thank you, Shubham. Thank you, Shubham. Thank you, students

Thank you so much, everyone. I really appreciate it. Thank you so much. Thank you. Okay

Yeah. Good night, everyone. Thank you, Rohit Jeev. Bye