How I Would Learn Data Science in 2022 (If I Could Start Over)

hi friends welcome back another video in today's  video i'm going to share how i would learn data   science if i were to start all over again a lot of  you have been requesting this video so i thought   i would share my thoughts and things that I've  learned that i now if i get to redo what i   would do differently if you have why some of my  previous videos you know that i am a self-taught   data scientist i actually did not go to school to  study data science my background is in business   and i had to make that transition through  self teaching through online courses   but there were definitely mistakes that i made  and now looking back let's say if i were to do   it all over again this is how i would do it and if  you find with information in today's video helpful   give it a thumbs up and subscribe to my channel if  you want to hear more about data science tech lab   cell and everything else in between so the first  thing i would do which i think i made a mistake on   is i just jumped in i heard the word data  science the sexiest job and i just jumped in like   data science is a huge job family there are dozens  of jobs within the data science and honestly i had   no idea about this and to me the jaw family  that was thrown at me or jaw family that i   was introduced to was a data scientist generalist  role so when i jumped into learning data science   that was the role i targeted but eventually  i realized that there's actually a lot more   options that i could have considered if i were to  start all over again for you who are just trying   to figure out who want to get into data science  and you're just trying to figure out your path   what i would want you to do is study the  data science field start understanding   what are the different job families within it  what it takes what is the type of work that   you would need to do to basically whatever  the type of work that you will be doing as   in that role and then align it to your interest  and your skill set figure out where you're going   to be good at and figure out where you're going to  enjoy working and then follow that path so that's   the number one thing that i would say is research  spend a lot of time studying online reading   talking to people who already work in that role to  get a better idea so that's number one thing that   i would do instead of jumping into one role that i  heard that i have heard of i would actually take a   step back and try to see like within data science  what are my options and what do i need to do   to go from point a to point b but you would need  to do a lot of research to figure out that point b   before you start creating a plan the second step  i would suggest is let's say you have picked data   scientist generalist as your role that you want  to target so the second step that i would suggest   we're not done researching yet we're going to do  a lot more research so look at the data scientist   generalist role and read the job descriptions at  different companies so study the role at different   companies and let's say if you have an ideal  company study the role study the job description   for that role and try to understand what are the  requirements that you would need to kind of check   to be able to qualify for that role and also a lot  of people don't do this go on linkedin if you have   already that role that description go on linkedin  and study the profiles of people who are already   working in that role because that's going to give  you a really good idea what are the things that   that person did that you need to do to basically  be able to be qualified per se so yeah that's like   my second thing that i wish i had done i actually  did it later on but that was i think that was too   late so if i were to start all over again i will  study the data science job family i would also   once i have narrowed down the role that i want  to focus on i would study the role and read the   job description at different companies the third  thing i would do is there's a lot of advice that   is out there that tells you to start with coding i  would not do that i would actually let's say if i   pick data scientist journalists as my role i would  not pick coding as the first thing i need to learn   i would pick the theoretical concepts when it  comes to statistics when it comes to machine   learning when it comes to business understanding  and i would have a very solid understanding of all   those things before i actually get to the coding  part i personally i think learning to code is   great but that's just remember that coding is just  a way is a tool to apply data science it's not the   data science itself there's a lot of people who  know python but they are not data scientists so   think of language the picking the coding language  as a tool to apply data science but you need to   understand data science conceptually understand  the fundamentals before you get to the next stop   because if you jump into coding the first  thing you are actually skipping a step and   you're jumping into the coding uh piece of it  where i would start is statistics like get really   really good at statistics because no matter which  role you are in you are statistics is part of   basically all the roles within data science so  you cannot go wrong learning statistics and if   you need more detail on this topic like i did one  of the videos where i talked about why learning   statistics is mo very important as the first step  to learning data science i would recommend you to   watch that video i have done a detailed video on  this if you haven't watched it i suggest you to go   watch that video where i talk about why statistics  learning statistics is important for data science   let's say you have done learning the theoreticals  you understand data science conceptually very very   well now is the time to apply the concept that  you learn on data that you're working with so   here i would suggest to start learning to code  so see the steps that we did so step number one   was researching the job family because data  science is a huge umbrella second researching   the role once you have narrowed it down third  getting really good at the concepts and um   understanding the fundamentals and theoretics  and then the fourth step is when you actually   start doing the coding start learning the  coding and for coding languages i don't think   there is any wrong choices but for me i from my  personal experience i actually started with sql   and r and um looking back i wish i had started  with sql and python i also did a video on this on   why which language i would choose python versus r  i would definitely choose python and you can watch   this video where i go into a lot more detail and  give real-world examples of why i pick python over   r but um i would say like if you are trying  to decide like python is a good choice so   pick sql python sql is a must no matter which  role you are in the data science i think a lot   of the people have to make a decision do you want  to learn python or do you want to learn r if you   have the option to choose i would suggest picking  sql and python now that being said sql nr is also   pretty good so i don't think any of these choices  are wrong it all comes down to personal preference   and remember if you have done the research  in step two where you understood the role   and you understood the job requirement that will  actually help you decide what language you need   to pick and what area you need to focus on that's  why researching is very very important piece of   creating your plan to learn data science  all right so now let's say you have   done your groundwork you have done the  learning fundamentals as well as coding   now it's time to do the projects so now  it's time to take everything you learn   and turn it into a project and kaggle is  actually a great place to start where you can   look at other people's projects and you can look  at what other people have done and there's a lot   of data available on kaggle so that's a great  place to start to kind of figure out how you   can build your project portfolio and then build  your project portfolio i'm planning to do another   detailed video in terms of like how to build your  portfolio but just know that if you have done the   research on step two you actually kind of know  what kind of project you need to do to be able   to stand out for that role i'm recommending that  you get an idea what type of projects will help   you qualify for that role or for those roles that  you are targeting so that's where you need to do   another connection between the project work that  you do and the project the the job requirement   and people in the projects that other people have  worked on who currently work in those roles so   do a lot of projects build your data science  project portfolio where you are able to show   what you know and you're able to apply the  concepts that you learn the theoreticals   as well as you're able to do the hands-on work to  show that you actually know what you're doing and   remember when you are trying to work on those  projects remember that start with the problem   and end with the impact and whatever you do in the  middle make sure to kind of highlight that as well   but i see a lot of people miss the problem  and the impact part and they only focus on   let's say if you work build a decision tree they  will be talking about like i build a decision tree   to do this but why did you do that what was the  impact of it so make sure like to follow that   framework start with the problem and end with the  impact and in the middle you can talk about like   how you solve that problem to get to the impact  part so those were the five first five things   that i would do if i were to learn data science  again so let's say you have done that there is   one last step that i think a lot of people miss  because a lot of people start thinking that   you have done the work and i made the same mistake  that you done the study you did the projects and   you built a portfolio now you should get the job  that's actually not true you need to now prepare   for the interview i'm sure you have heard that  interviews is a completely different monster in   itself where it doesn't really reflect like i'm  going to say this it doesn't really reflect how   well you know it actually reflects how well you  know in a limited time frame under in a pressure   setting so let me say that again the interview  process is trying to see how much you know in a   limited time frame where you're given a problem  and you're able to solve in a limited time frame   under in a pressure setting that's completely  different thing from the day-to-day work to the   interview and i know that's like a completely  different topic this is a different topic for   another video but just know that just doing your  data projects doing the learning is not enough to   actually land the job now you have to put in the  work to start interview preparing do the hands-on   coding prep where you are solving problems  in a limited time frame and do a lot of   mock interviews brush up on your fundamentals the  things that you learn brush up on that do a lot of   mock interviews with your friends and try to see  where you are good at and where you can improve   and then iterate on that all right so those were  the six steps that i would do if i were to start   learning data science again so let me know in  comments what has been your process for those of   you who did do the self teaching who did follow  the self teaching route what was your process   do any of these steps surprise you or resonate  with you let me know in the comments and with   that thank you so much for watching i hope you're  having a beautiful day i'll talk to you later bye

Channel: Sundas Khalid

Views: 108,891

Keywords: data science, data scientist, self-taugh data scientist, tech jobs, amazon, women in tech, big tech, google, Machine Learning, Python, data science projects, data science tutorials, data science jobs, data science day in the life, AI, data science 2022, big data, data analyst, data analytics, business analyst, sundas khalid, sundas, women of color in tech, people of color, people of color in tech, data science at google, data science roadmap 2022, data science for beginners

Id: 65yXoFc5stI

Length: 11min 22sec (682 seconds)

Published: Sat Jan 01 2022