How I Would Learn Data Science in 2021 (What Has Changed?)

my video about how i learned data science and extremely popular and i wanted to expand on that and talk about how i would learn data science if i had to start again from absolutely ground zero wow i had a pretty terrible fashion sense back then so by far my most popular video is the one i made about how i would learn data science if i had to start over wow i made this at the beginning of this year and i thought it'd be fun to remake this video and include some of the things that i've picked up since then as many of you know i started the 66 days of data to relearn data science from the very beginning i found out a tremendous amount about how i personally consume information i also recently saw an interesting critique of my last how i would learn data science video by andrew mo that's worth checking out after you watch this video these really got me thinking about my learning journey in a new light so much so that i wanted to make another video on this topic first let's talk about the technical progression and then we'll talk about the steps that you can take to actually get this done okay now what technical steps would i take if i were doing this over again in 2021 first i believe that you should start by learning the programming i find that with this skill set you can start building things which allows you to apply almost everything that you learn going forward quick example let's say i wanted to better understand how gradient descent works if i learn the programming first i can find code to build it and see it work on real data for me that makes it far easier to understand than just staring at a formula next i recommend getting familiar with statistics and data science algorithms for this i maintain that you should do just enough to get started with working on real projects this isn't to say that the statistics and the algorithms aren't important again i think for most people it helps to get your hands dirty with these first when you go back and take a look at the theory you never know it might just click for free resources kaggle.com is great i also have a discount code for 365 data science below if you're looking for something more robust they're in the process of switching to a new learning platform and honestly it looks pretty awesome you'll likely ask when am i ready to move to the next step whether it's after learning coding or after learning some of the basic statistics honestly most people will never feel like they're ready to move on you should keep going in spite of this for programming you should have a basic understanding of variables of loops functions and how to deal with packages for statistics and data science algorithms you should understand the difference between regression classification and clustering you should also be able to manipulate data with pandas and train a model with scikit-learn again theory is extremely important but we should go back and learn it as we use the individual algorithms more now my third phase is code review this is something that i didn't stress enough in my last video go on kaggle and clone some of the workbooks run the cells rewrite all of the code take notes on what you don't understand this will get you comfortable with the packages the syntax and everything else in between it also helps you to start building intuition now it's time to build projects get obsessed with these do as many high quality projects as you can these will compound your learning and i'll help you to build a portfolio that will land you a job one day i recommend doing at least one project on the three main types of data science problems the first is classification the second is clustering and the third is regression this is by far the best way to actually ingrain the skills in my own life i learned about five times more from the projects and the work experience that i got than any course that i ever took an important part of building projects is also sharing them getting relevant feedback and being held accountable are things that can supercharge your learning on that note let me know in the comments section below about what projects you're planning to work on in 2021 i love hearing about these different things and they really get me excited to do my own projects before we move on i have a couple very important notes to make first these steps don't necessarily have to be done sequentially they'll most definitely bleed into each other second you're never done learning these things it's a continuing process you never master python or master an algorithm you're constantly reminding yourself and refreshing your knowledge as i mentioned before you'll also likely never feel ready and you just have to constantly push yourself out of your comfort zone finally for most people including myself this is not easy you get stuck you wrestle with things you go on stack overflow you get frustrated you start pulling your hair out this is part of the career while it sucks it also can be one of the most rewarding parts of this profession there are a few better feelings than getting a solution to work that you've been struggling with for the past couple hours oh wow uh looks like i got just a little serious there with you but let's move on to the things that i personally learned this year about my own learning so the first thing that i change about how i went about learning data science before would be to start with a plan this time for my 66 days of data challenge i thought i'd just go back through learning some of the things that that i thought were interesting and brush up on resources that i used in the past i'd also mix in some projects here and there honestly this is pretty bad idea i was constantly fighting for what i should work on next and how to integrate it into my my new journey i recommend following some structure through this this next learning progression you can use an online program like cackle.com or 365 data science like i mentioned before or you can create your own curriculum most universities have their curriculum online for free so you could see how they're building out their coursework and you could build one around your interests as well as some of the structure that they have existing there yes i'm in the bathroom i'm running out of places to film the next thing that i'd recommend is using multiple resources it can be very frustrating to not understand something i even personally have issues with self-doubt when this happens sometimes though it's not my fault i've found that the same concept explained in a different way can produce incredible light bulb moments for me if you're struggling with something try looking at four to five different resources you might be surprised at how this little change can make things far easier to digest i also recommend this approach for choosing paid courses most of the content is very similar i believe most of the resources are very good but if you resonate more with the structure of one teaching platform or with one individual teacher that is definitely a reason to choose one course over another you know if you don't like hearing me speak there's no reason to to buy my course and hear me talk for four hours you'd be torturing yourself the third thing that i'd recommend is to make learning a habit as many of you know i'll be starting the 66 days of data journey again at the beginning of 2021 i along with over 2 000 other people have found tremendous learning growth through this initiative simply learning something every day has a huge compounding effect on your trajectory as a data scientist the last thing i would do to learn data science in 2021 is to engage with communities more i've loved learning alongside others on my youtube channel the 66 days of data discord and on linkedin if you want to engage with this video you can always hit the like button and subscribe if you haven't already people have been far friendlier than i would have expected when answering my questions and giving me constructive feedback and i encourage you to be an active member in these communities obviously you should do your homework first before engaging but my rule of thumb is to not ask a question if you can easily find the answer on google if you can't find an answer definitely get involved answer questions ask your own questions the more you give the more you receive i think 2021 is going to be an awesome year for data science and for your own learning journey and i'm excited to kick it off by again participating in my own 66 days of data challenge i really hope that you'll join me we have some awesome companies like hp nvidia 365 data science interview query a lot of these companies are giving away some free stuff they're going to be heavily involved and i think engaging in this community can also pay dividends on your end being able to have access to a bunch of really cool and free resources thank you so much for watching and good luck on your data science journey be sure to check out these videos next also feel free to take a look in the description i have a ton more resources and i also for every video i make give a description of all the main points in there so if you really don't like hearing me speak i also have the the other option of being able to read all the content there

Channel: Ken Jee

Views: 113,659

Rating: 4.9611297 out of 5

Keywords: Data Science, Ken Jee, Machine Learning, How I Would Learn Data Science, How to Data Science, Data Science Learning, Data Science Tutorial, Data Science Journey, Data science timeline, data science career path, how to learn data science, how to learn machine learning, data science education, python, learning python for data science, data science career change, data science course, data science from zero, data science 2021, How I would learn data science in 2021

Id: 41Clrh6nv1s

Length: 9min 8sec (548 seconds)

Published: Sat Dec 19 2020