Akansha Yadav explains what machine learning is and how it differs from traditional programming. She covers everyday examples like Netflix recommendations and loan approval scoring, then maps out the landscape: artificial intelligence, machine learning, and deep learning. The video introduces supervised, unsupervised, and reinforcement learning before surveying applications from computer vision to autonomous vehicles.
Chapters
Ladies and gentlemen, hello everyone, welcome back to my YouTube channel. In today's video, we are going to see what machine learning is and how it works in the real world. So let's get back on the screen and see. So we will discuss machine learning in depth throughout the series. This is a complete series where I will be sharing multiple things about machine learning from basics to advanced, and also the projects of machine learning, that what you can build and how you can build. So let's first see what machine learning is. Basically, machine learning is the subfield of the computer science that gives the computer the ability to learn without being explicitly programmed. So earlier what was happening is programmers were there who were typing the code, and on the basis of that code which they were typing, the computer was able to, uh, give us the output. But that was limited to the program that has been programmed by the programmer. But now it's the time where machine learning is taking a charge, where it is making the computers able to understand, able to learn without being completely programmed. So how it works is basically we provide the data to the computer and we ask our data— our computer to learn from that data, and whenever any question is asked, it has to respond on the basis of the learning that they have made on the basis— on, on the data. Basically, let's first see what is the basis of— basics of AI. Machine learning follows the same process, just like a 4-year-old child who uses to— who used to learn, you know, basically used to learn, understand, and differentiate animals. So machine learning algorithm inspired by the human learning process, which learns from the data and allows the computer to find the hidden insights. So basically, at large, data has been provided to the computer Computer learns from that data and computer will find the hidden insights, hidden patterns from that data and will learn it from that. And whatever the computer has learned, it will return us back whenever any question is asked. How do you think Netflix and Amazon recommends videos or movies or TV shows to the user? This is the real-world example of machine learning, that how machine learning works. They use this machine learning to produce suggestions that you might enjoy. This is similar to how your friends might recommend a television show to you based on their knowledge of types of show you like to watch. Suppose me and me and my friend are discussing about something which is full of drama. So my friend know that I know I, I love to watch something which is full of drama. So my friend will suggest me something which is full of drama because my friend knows me very well. Same way, the data has been provided to the computer, and computers now know that I am the user who likes to watch the drama movies on the basis of the history that I have been watching on Amazon. I have been watching on Amazon videos or Netflix and all. So my machine, that machine learning will help me, uh, it will basically throw those content onto the screen for me that this is the content which you may like. That's how machine learning works. You can think of another thing that how do you think bank makes a decision when approving a loan application? Loan application is something that there are two parties with who do not know each other, right? So how they can make the decision, how they can come on to the point is they uses machine learning to predict each applicant's default probability and then approve or refuse the loan application based on that probability. So that's how machine learning works. Machine learning will take the details of that applicant and on the basis of that details it will try to predict, it will try to evaluate the probability that whether that person is a defaulter or not on the basis of the transaction history or the other details. Then on the basis of that, they will approve or refuse the loan applications. There are multiple telecommunication companies which uses the customer's demographic data to segment or predict if they will unsubscribe from their company from the next month or not. So this is where machine learning comes into the picture. First is the artificial intelligence, which is— which says that it is the ability of a machine to imitate the human intelligence. It is trying to mimic the human intelligence. Within the artificial intelligence, there is a domain called machine learning. Machine learning is a part of artificial intelligence which says that it is basically an algorithm to incorporate the intelligence into the machine by automatically learning from the data. As I told you, that the data has been provided to the computer and computer learns from their data. We do not program particularly to the computer, we just ask our computer to learn from the data which has been provided. Another is the deep learning, which is also the subpart of the machine learning, which mimics the human brains to basically mimic the intelligence into the machine. Machine learning and deep learning are differ— like bit different, but we will discuss it in detail about this in the next video. But as of now, you can consider machine learning as the ability, as an algorithm, as the model which can help the machine to learn from the data by automatically learning. There are different type of machine learning models. One is supervised, unsupervised, reinforcement learning. You do not have to worry much about it because we will be discussing each of these things in the upcoming videos. Of this series. Let's discuss the machine learning applications in the various domain. There are multiple applications in the current world, like computer vision, where there are multiple areas where image recognition is taking place. And Google Photos, or in the Apple Cloud also, image recognition is there. It can detect the objects, it can detect the persons, it can do the facial recognition as well. Another domain is the natural language processing, just like language trans— translation, which Google Translator can do. Another is the sentiment analysis. Sentiment analysis is something which is done— suppose I have the multiple reviews of the product, and those reviews are analyzed that whether they are positive reviews or they are negative reviews. Another is the chatbots like ChatGPT and all. In the healthcare also, machine learning is taking place in the diagnosis of the disease or the personalized medication to that particular patient, or the drug discovery. And the finance, as I already told you, that there is something called stock market, there is fraud detection, risk assessment, algorithm trading, and many more financial things are there in which machine learning is in work. Autonomous vehicles are also there, just like a Tesla one, which is driving on their own on the basis of the map that they have been provided. It can do the object detection, it can do the path planning and the decision making as well. So these are the various domains where machine learning comes into the picture. In the next video, we will be discussing more in detail about artificial intelligence, machine learning, deep learning, and what are the different types of machine learning models. So if you like this video, don't forget to hit the subscribe button because we are going to discuss more about machine learning. And you can comment down your views if you like this video or if you don't, or what you wanted to add on in these videos. That would be really helpful for me to make the informative videos for you.
