How to switch your career to Data Science.

Key Takeaways

hello all today you'll be seeing will be understanding how we can transition our career towards data science this question was asked by many of my subscribers so I want to take up this video and explain you how you can transition your career from any domain towards data science or make sure you watch this video till the end of the section but I'm going to discuss a lot of things into this and make sure you understand these things and then you can apply your is the same thing I have got a teaching experience of more than three plus years so I have seen people from many backgrounds many different backgrounds like HR medical education medical healthcare people from financial domain people from computer vision domain images videos all this kind of different different domains people have come from and they have been successfully able to transition their career towards data sides so I'm just going to discuss how do we start with it with our transition part and what are things we need to take care of and how by learning with joint technology will be able to achieve this data science career now to begin with always understand that case um nowadays people are from different different domains when I first of all I'll just discuss about the people who are from programming languages to me right so programming language is domain so I've seen people who are from domain of dotnet Java you know and then people have been working in Python in other programming languages like our and and they have been also working with along with dotnet they are working in a skill server domain basically in the database domain here in Java we have also l and people have also been working in oracle domain you know as a pl/sql developer so all this curve so you can see that the most of the programmers nowadays are usually working no and all this different kind of programming languages and they may be other programming languages let off I would like to include it but I'm not included right now just to show you an example you know these people what they are like what I have seen is that people are trying to move from programming domain who they are working in dotnet and Java into data science right now what they do they need to do in order to switch the courier towards data science now see the programming stuff that they have been learning in this dotnet domain you know or Java domain of Athan domain that programming stuff is very very important because how many years of experience they may be having like 10 years experience or 5 years experience or 2 years experience no matter if they are also a fresher and if they won't know one of the programming languages not they will be very they will be able to switch they carrier very easily because since they have the programming background in order to learn Python so basically in the data science domain when we talk about machine learning and ah we basically have two very very important or very very nice programming language one is Python and the other one is our yes with Java and C++ also we can do the data science basically we can apply machine learning a lot but in Java and shape C++ we don't have that many libraries like we have in Python and hard to apply the machine learning algorithm so there are some libraries which we call it as scikit-learn you know and there are a lot of deep learning libraries like tensorflow you know or Karass which are very very easily supportable in python and people extensively using it making the work easier so basically it is not like you cannot implement with Java or C++ you can definitely implement it but just understand that if you are writing 200 lines of code in Java or C++ to implement a machine learning algorithm that can hardly done I hardly be done in 10 different lines of code in Python so so people who are from programming background it is definitely easy for them how to switch them switch to the carrier the first of all they need to understand they need to know how to do the programming in Python and I tell you guys I feel python is the most easiest programming language because I was also a.net developer as my background like initially one to one and a half years I have actually worked in dotnet domain and then I switched my career in to data science so I will be the right person to tell you for the programming people how you can actually switch it so when you are having some programming background basically it is very easy to learn Python you know Python is just a simple language there are lot of tutorial sites that are freely available but the main thing is to understand that how do you apply Python in data science in any machine learning use cases or deep learning use cases so after learning Python right you need to understand some more things some more topics that you need to understand is first of all is very important concept for less statistics the reason you need to understand statistics guys because in machine learning and deep learning right machine learning and deep learning this this both the concepts are completely based on statistics in the backend there are a lot of mathematics that will be involved into this lot of mathematical calculations and that term in whole will be basically called as you know statistics is in shock so there is a lot of mathematics involved is lot of linear algebra involved is lot of probability involved again this statistics becomes very very easy in Python because the reason is that they are lot of libraries lot of libraries available supportable with Python programming language and in our programming languages so in Python you basically have a library called scikit-learn in the sky could learn you basically have a lot of pre-processing libraries that involve different different kind of statistics apart from that in Python you also have some people as stats model so stack models we just start models just and models library also includes lot of statistics concepts for ittogether the next the next thing is basically to understand how how what our topics are actually involved in statistics now see you guys we will be applying machine learning machine learning is the concept where we are making a machine to learn based on the data that we have so start is it inverse a lot of statistical tools within it because we need to understand how the data is actually distributed so distribution of the data is very very important now if you go to my playlist and see I will be going to my towards my playlist and showing you two that have created different different playlists for all the sections you know first we need to start with Python then we need to understand some concepts of statistics where we will be understanding what is the data actually how the distribution of the data is making and back in the use case and how we'll be able to solve that particular data by seeing a kind of distribution in that particular date so all those things are basically covered in my playlist so that is a motion but anything that you need to understand so first step we begin with Python then we begin with statistics you know then apart from that after you understand statistics that basically means you have understood how the data is actually distributed after that you basically do something called as feature engineering C feature engineering is again a very important step space feature engineering helps you to understand about the future you know whether your feature whether your data set so data set when you say data set it has lot of columns right and each column specifies one feature so what this actually speech feature is specifying do you need to do any pre-processing to change this particular feature that is where the future engineering comes into existence now again feature engineering for the future engineer you need to understand some of the library's very very clearly you need to learn this libraries basically you need to know how to program with respect to this libraries that is numpy and pandas see you guys if you understand numpy and pandas most of the feature engineering work will be very very easily done so you need to when you're learning Python right you should definitely learn numpy and pandas which are one of the most important life is that are used in Python C if you are able to understand pythons very nicely along with the practice you know that that is the reason what that is the case where you will be able to excel in the feature engineering process most of the time in a data science project goes into this particular feature engineering step so basically 30 percent of your overall time goes into feature engineering because this is the step where you need to understand about each and every features whether those features are actually required so in this number in finder so that is also I'm going to consider this as a part of pythons now in Python what else you should need to learn first of all you need to understand about the different different data structures that are present in Python that is like list number list dissing Ares arrays you know sets tuple so all this different kind of additional I mean data structures you need to understand apart from that you need to understand some of the libraries that are available in numpy pandas and there is one more library which is fine as matplotlib matplotlib is basically used for the visualization purpose and that the next library that you usually use is something called SC bond now see bond is also very important library's days because Seaborn helps you to plot or visualize the data based on the statistic concepts that you are trying to apply so it is very very important after you understand Python with numpy pandas matplotlib and c1 c1 as i said that c bond is basically used for visualization of the data applying statistics into concepts so a lot of statistics will also come inside c bond which will be able from which you will be able to analyze your data properly now after understanding all this trust again this is all included in python days in python you need to also understand how to create classes you know classes functions you need to understand how inheritance work so that actually helps you to reuse a code okay so python part is basically completed one more topic you need to understand in Python you see something called as exception handling because whenever you are creating an a whole scale application and definitely exception handling will also be helpful for you so after understanding all these things guys then you move into machine learning then the next step is basically to understand what is machine learning what is ml you know and what is DL what is the difference between ml DL what is artificial intelligence that is al okay in short if you want to define LD L animal basically you can just consider a small graph you know al is basically the application that you are creating within this there is a subset which is called as ml and within this there is another subset which is called as DL so machine learning and deep learning I have already explained everything in my playlist I'll go through the playlist at the last to make you understand what all things are present in through this and how you start learning through it so after you understand this you need to understand ml ml now whenever you do machine learning guys the first step is again data gathering you don't have to worry much about the guitar you add link because data gathering is basically done by a data engineer okay big data person or person who is working for the back end but sometimes you a data scientist may also have to work in data gathering and that time is basically called as a full stack data scientist so after data gathering what you have to do is that you have to basically do something called a data analysis or I will basically call this as exploratory exploratory data analysis so this step is then done then you do feature engineering you know feature engineering after applying feature engineering you then apply your machine learning machine learning is basically you will be applying algorithms they have various algorithms that are already available in Skype Hitler you know like linear regression you know basically for two different kind of problems one is regression and the other one is basically called as classification so two different kind of use cases you will be basically getting so you have different different libraries like linear regression you have decision trees you know you have random forest okay you may be having eggs eBoost you you are having a zero sorry you not may be having you are having eggs abuse then apart from that you also have something for a logistic regression you know k-nearest neighbor KNM then you have something called as name bias this you will you'll also be learning about natural language processing see the best thing about my channel is that I have actually created videos on every of this particular topics you know around 80 videos are actually present which you can use it for yourself so this is how you actually learn about machine learning in machine learning you need to understand about this particular libraries or sorry about this particular algorithms and then whatever data you have actually got after doing the feature engineering and pass it to your machine learning model you know and then your model will be start giving you new outputs for the new kind of input data that you provide here so this is basically like a black box all together and you will be able to get this particular data now you can see that from this particular background and this is for the programming background again guys for non programming people right non-programming people because I have trained some people who are in HR domain you know HR domain in healthcare domain you know what I faced is that they were facing the problem to understand programming language mix because they don't have any hands-on experience to understand programming language so programming languages because they don't have any idea about C++ Python sorry C++ Python Java or.net so the most of the time that date took was to understand how they can program in Python you know so suppose if you are trying to transition your career or if a non programming person from a non programming background he was trying to transition if they're here most of the time he'll be understanding he'll be trying to you know implement Python you try to learn Python you know and if you are planning to do it in six months so basically half of the time you basically be gone in understanding Python how to program in five now in Python I am NOT just saying you know just just to you you are learning some data structures and I'm talking about all the libraries that I discussed just like numpy pandas you know and matplotlib matplotlib and Seaborn all these concepts so all these concepts were basically they took time but after they grabs that particular power you know then they were very very easily able to transit their carrier towards data science because they had that domain knowledge I just said about what is going in what is over here now another question that I usually get is that some of my subscribers know my friend says that Krishna 10 plus years of experience or I am 20 plus years of experience can I switch and I was working in so-and-so domain okay domain means basically any programming languages any specific domain that working can I transit my carrier towards data science that is the next question that you they usually ask I say yes a big big gears a big big yes I say them yes because there is a reason why I said yes oh you need to understand that guys whichever domain you work in whichever programming languages you work in you have some domain knowledge you know as a data scientist a data scientist needs to do the work of both programmer you know communicator then as a data analyst so he has to do the work in all the section when I say he should be a communicator that basically means that he need to talk with the client to understand that particularly useless you need to talk with the client like what kind of data he's actually requiring it and after that getting that data I'll analyze that particular data so he's also doing the work of a data analyst and finally he is working as a programmer to create the machine learning model and it is actually participating in the deployment process so guys any person who is more than 10 plus years of experience so all the all the experience that they have in a specific domain right that will be very very useful because he'll be the person who knows how to communicate things who knows how to handle the project so all those experience will not go waste guys he just need to understand oh he just need to practice some Python programming language and understand some of the libraries that are available in Python and then he can easily easily change its courier to any other background ok and yes for the freshers it is very important that they can transit the carrier very easily very quickly because they will be in the mood of study mode you know and they will be they'll be having some knowledge of programming languages and obviously nowadays companies are looking for more more people more freshers because each and every company is creating different departments on a IML or where they are solving a different different kind of useless now from this particular diagram there's you can see that as your background is a data engineer you can basically could easily transit to beta analyst business on a professional a predictive analysis analytic profession this this basically means that this whole everything can be involving with data science you know he may be a ml engineer he can become a DVD burning engineer the next thing is that nowadays lot of MBA background people who are doing and I excite they are also they also will also be able to change their career to as a data scientist apart from that people who are working who are actually doing the education in statistics number of science engineering students all those things even mechanical background students can also change I've seen people who are changing the career in to data science I've also had some of my friends who are from mechanic background to switch the carrier to its data science yes now this is what is the whole process of doing it guys I've told you all the steps let me revise it once again so first of all is that you need to learn a programming languages either it will be Python or are you know then you understand the techniques techniques of the data science lifecycle and about the data science lifecycle I have already created a video or in my playlist but again I will show you all the playlists so after you understand the techniques of the data science lifecycle the next thing is that how do you deploy the models how do you deploy the machine learning models now the feature engineering all comes inside this particular life cycle itself all all about the data data clip crossing everything will come in the second step in shock so let me go ahead and show it to you guys like how what all playlists you have to look after that and I actually created all the playlists so you don't have to worry about much of over it so here it is so so far let me introduce myself my name is Krishna guys I work as a lead data scientist I have around seven point five years of experiences in data science and no seven point five I have somewhere around five years of experience in data science and one point five eight years of experience in dotnet so here it is my first the playlist that I have actually created of around 36 videos I have introduced what is machine learning from the scratch then you will be seeing what is supervised machine learning anaconda installation how to do do it since it is an open source just follow this playlist to understand machine learning but before that make sure you know Python a bit of Python you know then I have implemented each and every machine learning algorithm like simple linear regression polynomial regression then we have logistic regression support vector machine k-means clustering - plus say make G boost or PI spark tutorial for beginners principal component analysis everything is that you see you have also created about time series data exploratory data analysis so part of each engineering then you also have decision trees everything like that now this was with respect to machine learning if you are interested in understanding of statistics then I have also created a different playlist from statistics and I will be continually continuously uploading these videos statistics though playlists - now you can see that all this particular like population was a sample what is Gaussian distribution these are basically the data distribution that usually coexist in this particular world different kind of distribution that we have is basically called as Gaussian distribution normal distribution log normal distribution then you also I also have created a playlist from feature engineering you can see different difference each engineering playlist is there with respect to national language processing also have included over a very it will help you to P process your text data now all this playlist URL will be given in the description of this particular video and apart from that after after feature engine and playlist is also separately given to you then finally a lucky one I've created 30 videos on data science it will be questioned so that you will be able to you know attract any interviews after you learn all these concepts and here also we have around 30 videos which will help you to you know transit your carrier towards data science so once you learn each and everything guys suppose if your target is around six months make sure you do practice for you do learning for three to four two to three hours daily take out your time yes you can definitely do it anyone can definitely do it but you need to know the life cycle of the data science project and everything guys make sure you learn this and if you like this particular video please do subscribe the channel share with all your friends who want to change the carry on towards data science yes it is now high time to change the carrier or because you have a lot of work that is going to come up many companies in you do the exponential growth of the data are planning to switch the carrier towards data science or that's all I hope you like this particular video and I love see you in the next video plot less all thank you one and all I'll see you in the next video thank you bye