School Of Basics | What is Machine Learning | Step by Step from Scratch

Hello and welcome to School of basics I am Raghav and today we have a very interesting topic to learn that is machine learning now do not worry if you have never understood what exactly is machine learning I am going to start from very scratch and by the end of this video you will have a complete understanding of the basics of machine learning let us take some examples to understand so that before I tell you the official definition of machine learning you will actually be able to understand what exactly it is and then you will be easily able to get the official definition think of a time when you had to plan a holiday now whenever you are planning a holiday you have to predict your expenses how much you will have to spend and all the expenses you will have to predict now it may include the cost of travel suppose you are going through flights you will have to check the flight rate you will check what are the what will be the cost of hotels at that particular time you will have to predict for the expenses for all the activities shopping and the food all the things you will have to predict and then you can calculate and then you can see what will be your expenses now to predict these expenses you can do a lot of analysis and get information from different sources for example you can go to the websites then you can search for the flight ticket prices at that particular time you can search for the hotel prices at that part of the season the activities that you have to do the prices of those activities the food shopping etc you can get information from Internet online websites you can also look at some reviews and take information from people who have earlier visited the place or you can take help from your friends who have all gone there and they can provide you information and data about the expenses and all other things so using all this data and information you can predict your expenses now machine learning works in the same way we provide data and information to the machine as much data as possible and looking at that data and all the information that is available it can take some decisions or can do some predictions if I take a very Real world example on your email suppose you got some email and you mark it as a spam and you did it two or three times every time you got that particular email or a similar kind of email you mark it as a spam now after sometimes your system will learn that this particular email or these kind of emails coming from some particular sender or having some a specific message of having some specific body this is a spam and the next time it will take a decision it will predict and take a decision that this is a spam email and it will market as Spam without you having to do it manually this is how machines learn you have provided data and information by marking that email as Spam two or three times and the fourth time it has learnt and now it can take a decision on itself so this in very very simple words what machine learning is now if I go to the official definition it says machine learning is the study of computer algorithms that improve with experience so as we keep on giving more data and as it keeps on doing some predictions and decisions it keeps on learning at it it keeps on improving with every experience and this machine learning algorithms build a mathematical model based on sample data we also called it has training data and based on that data it can make predictions and decisions without being programmed to do so so we have not programmed for doing some particular action but based on the data and information it can take a decision and predict and then do some actions without we having to code for that particular action and this process of using data and then doing prediction is called as regression in machine learning with this very very basic knowledge let us go forward and learn about the components of machine learning now talking about the components of machine learning there are three important components the first one is data or data set a collection of data is called as data set and data is very important in machine learning as we have seen earlier we need to provide a lot of data to machine so that based on that data they can do some predictions and then they can do calculations and then take decisions and it is important to provide data and the more data you provide and the wide variety of data you provide the better it will be the better will be the calculations the better will be the predictions and better will be the decisions so data is the first and very important component then we have features Now features are individual measurable properties for example if you go back to our example of calculating the cost of our holiday in this case features can be the number of days the airline services the type of hotels like 3 4 5 star hotels you want to book all this will come as features and features can be feature can be numeric it can be a string for it can even be graphical so this is feature and second component of machine learning and then we have algorithms now algorithms are a set of instructions to solve a problem so the sequence or steps or instructions and algorithm uses features and data and based on the data we have and the features we have it runs some steps or sequence of instructions and based on algorithms we can actually do the calculations and then do predictions and can take a decision so machine can do these calculations prediction and take decisions based on these algorithms so these are the three components of machine learning now if I give you a very Real world example whenever you try to do some search on a e commerce application or if you have searched for a product on a eCommerce application or maybe you have searched for some flights for your holiday now whenever you do a browsing on the Internet and you are on a completely different web site still you will see ads related to your search that you have done in the past so it will show you ads based on the products you have searched or similar kind of more products or based on the flights have searched so the system can learn that you have done some search for this kind of things and based on the all these components it can now take decisions and show you the relevant ads so this is a very good real example of how machine learning works now with this knowledge let us see the different types or approaches of machine learning and before I go there let me also show you this some people are confused between machine learning and artificial Intelligence and deep learning so deep learning is a subset of machine learning and machine learning is a subset of artificial intelligence so machine learning uses deep learning and artificial intelligence uses machine learning now I will explain more in detail about these topics in a separate video but this is just for General information and now we can go and learn the different types or approaches of machine learning there are three approaches or three types of machine learning we have supervised learning and unsupervised learning and reinforcement learning now in supervised learning we find pattern based on both input and output data for example the decision to Mark a email as spam this is an example of supervised learning or to predict the winner or loser of a game or to predict whether a user should be given a loan or not all this will be a type of supervised learning know when we talk about unsupervised learning here we find the pattern based on only the input data and we are not having any output data now this is very useful when we do not actually know what we are looking for example visual recognition or putting the name to your friends photo on Facebook this is a type of unsupervised learning approach of machine learning or in case of Google photos Google can a Google photos can put a photo into a family album or office album based on the facial features or some other features on this will be a type of unsupervised learning then we have reinforcement learning and this is when a machine or your computer has to work in a dynamic environment we make use of reinforcement learning for example automatic traffic lights now in case the traffic Lights Are automated to change the signal or change the lights after a set duration that is actually not machine learning but if we have traffic lights which can check the conditions of the dynamic environment and see which lane has more traffic and it provides more Green Signal to that lane and seas and predict which lane has less traffic and can predict and then take a decision to provide more that signal to that Lane that will be a type of machine learning and the machine learning approach used here is reinforcement learning now if you have any confusion or doubts in the components of machine learning OR the types of machine learning do not worry the motive of this video was just to introduce and to explain the very basics of machine learning and if you are somebody who understands better with A story I have written a story on machine learning and you will find the link in the description of this video and I will meet you with a new topic in the next episode of school of basics thanks for watching

Channel: Automation Step by Step

Views: 2,059

Rating: 5 out of 5

Keywords: automation testing beginner tutorials, what is machine learning, difference between machine learning and artificial intelligence, components of machine learning, types of machine learning, machine learning for beginners, machine learning interview questions, machine learning basics, learn machine learning, machine learning from scratch, getting started with machine learning

Id: oDNT0OKftW8

Length: 9min 53sec (593 seconds)

Published: Mon Aug 17 2020