Thanks for watching everyone! Please leave some of your data science goals below so you can be held accountable by the community on this channel. Also, feel free to like and subscribe to help this channel grow! Watch this video on why I am starting data science over again next: ruclips.net/video/uXLnbdHMf8w/видео.html&ab_channel=KenJee ! Video on my #66DaysOfData Challenge: ruclips.net/video/qV_AlRwhI3I/видео.html&ab_channel=KenJee. Written version of this article: towardsdatascience.com/how-i-would-learn-data-science-if-i-had-to-start-over-f3bf0d27ca87 A few more links: My Favorite Free Data Science Courses: ruclips.net/video/Ip50cXvpWY4/видео.html&ab_channel=KenJee 365 Data Science - Courses ( 57% Annual Discount): 365datascience.pxf.io/P0jbBY (Affiliate Link)
01:20 - 1) Learn programming in python or R 01:40 - 2) Learn basic statistics for data science 01:58 - 3) Start projects as soon as you can (Kaggle) 03:28 - 4) Take notes on what algorithms and packages others use 03:46 - 5) Learn the source code for the algorithms and try to implement them from scratch 04:48 - 6) Work on more advanced projects where you collect your own data or use advanced concepts like deep learning, nlp, or computer vision.
Best advice from a data scientist: 1) nothing is what it seems. If you have yet to enter the data science and analyst field (still in school or studying), what you think a data scientist is day to day is wrong. So keep an open mind. We’re not writing algorithms on the windows of dorm rooms. Hint: 90% of our work is research and data preparation/cleaning. If you want to be a kick ass data scientist, learn how to interview data users. Coding is the easy part. 2) did I say be open minded? Good. Now tell yourself over and over and over, “almost no companies know what data science is”. And neither do you. So let me tell you what data science is: it’s the ability to understand and manipulate data and/or information. That’s it. Data science is not Python. If I give you an excel sheet of data and ask you to create a pivot that shows me counts and sums of those counts and you do that; you are being a data scientist. Your ability to code in a certain language does not = data science. I’m telling you this, cause being and looking this internet conceived data scientist part/person is a misconception and can kill your dream of ever growing into a great data scientist. Give yourself credit for what you know. And be confident that because you can analyze data in excel, you can also learn object oriented programming. 3) nothing is what it seems. Just cause a job title doesn’t say data scientist doesn’t mean your not going to do data science work. I’m a business intelligence developer. I manage data from Oracle, SQL Server and Cloud data lakes. When I sit down at my desk, I open Dbeaver, SQL Server, Visual Studio and Jupyter Notebook. All these IDE’s get used daily. That means I coding in 3 languages at a minimum daily. ONLY ONE of those languages is Python. My job title is NOT data scientist. My point in all my rubbish talk is that data science is a lot like the cloud. Over 10 years ago, the cloud concept became the rage. And yet most companies are still struggling to implement or even leverage the cloud. Data science is sadly going down that path too. Most companies and their managers are posting jobs to hire data scientists and as long as the job posting has a requirement of “must know Python”, it’s a data scientist job posting. However, 99% of the managers hiring data scientist don’t know what a data scientist is themselves. And those managers couldn’t tell you what a Jupyter notebook is. Or how to write/print “hello world” in Python. Yet they’re interviewing candidates for Python development. 😁 If and when you enter the so called data scientist career path, be open minded. This is a journey not a goal. Enjoy the trip. Be flexible. Knowledge trumps job titles. If you want to get paid the big bucks, be the most valuable player. That means work on everything sent your way. And show off your skills. Good luck!
this is a bit weird comment... you say your job is "90% of our work is research and data preparation/cleaning [...] That means I coding in 3 languages at a minimum daily. ONLY ONE of those languages is Python", and then "I’m a business intelligence developer [...] My job title is NOT data scientist". I think you are indeed not a data scientist, and your job is not a data scientist job. This sounds a lot more like a data engineering job, or as you said, business intelligence developer. I'm sure it is related to data science, but I feel it's not quite the same.
@@dhidhi1000 I’m actually a Lead Data Scientist now. This is a bit of an old post. I think where we get lost in translation, is that I don’t work for one of the FANG companies. My company doesn’t blow money on a team of 5, all developing in one language out of VS Code. I have a team of 3. We support one department effort. And that means we not only own the data, we build the data, we run all analytics on the data. If the department we support needs to know the impact of a business decision, we perform the data science analyses, produce findings and reports, and build out the automation/bots. And not bragging, just being truthful; pure data scientist and certificate data scientist don’t last on our team. They usually tap out, because they are looking for a assembly line like work environment. They want to do one thing. For instance, they want me to “data engineer” and then hand them perfect data they can apply analytics to. We don’t do that. We are ninjas on my team. So 3 languages at minimum. SQL, Python, Java. And you also need to have a robust understanding of business operations. Spending time with the VP is not unusual. We make data scientist Managers and Director’s in short order. Because we bring our data scientists into all aspects of the business. I love our setting. Never a dull moment.
This is the first video I've been able to sit through engaged in a genuinely interested and joyful way on the topic of data science. You're not gimmicky or salesy. Thank you so much!
I'm very young and I've been learning a lot of programming for many months. I had already started when I was 13, but nowadays I feel more enthusiastic than before. This kind of videos are really helpful and motivating. Thanks!
I was very interested in Machine learning and data Science. I would like to see all your videos. They may help me to achieve my goal. Please do many videos that encourage and guide the beginners. Thanks 😊
Love this! As a pragmatist I find it appalling how intimidating this field appears before you start diving in and getting your hands dirty. Hopefully people can start to understand that DS is human intuition applying math and computing as a tool but not a concept separate from real word problem solving.
The thing I really like about your video is the stress on getting stuck into projects. I am a few months into learning about Machine Learning, but it is only since I started focusing on actually doing projects that I actually feel like I am beginning to understand what is going on. Before, I was focusing on programming, algebra, statistics, and I felt like I was just going through the motions, but now that I started trying to do the projects its like I am waking up and finally learning. Wish I had gotten stuck into projects much sooner.
I've learning data science for about 3 months now. My biggest problem was one those you mentioned in the beginner mistakes video : studying too much theory, trying out too many different courses . Your videos are great , man. Really helpful. Thanks. It made me more confident about making mistakes and trying out new projects . Not just kaggle , other more difficult platforms too.
Thanks Ken for a great video on the roadmap of learning data science! The intro animation graphics looks awesome! Well said, accountability indeed helps to keep us motivated and pushing it to meet the goals we set. Being accountable for one another in this comment community is surely a great way to move forward in learning data science together. 😀
Here's a course Face Mask Detection Using Deep Learning . It's paid but it's worth it. khadymschool.thinkific.com/courses/data-science-hands-on-covid-19-face-mask-detection-cnn-open-cv
Hi everyone. I am graduating from college this year with an economics degree. I have about a year of experience working as a data analyst (MySQL, Stata, R, Python proficient). Since startup recruiting has been a bit tricky because of the virus, I want to use the next few months while I am unemployed and living at home to become more proficient in higher ML techniques.
My goal is to create a chatbot for my coaching business that will kick off on Udemy by next April. I just completed a month long Python boot camp and can now begin the real conversation design. My roadmap is in a paper I presented last September in Durban, SA - just as motivation to keep me at it. I’m a cybersecurity analyst and acquired my master’s degree in Linguistics at UGA in 2017. I guess I’m saying all this to force myself to stay on track and change careers to machine learning. No one around me (including hubby, kids or friends) is interested in machine learning enough to make this their second (non-paying) job, the way I have 👀-> as of now. Between home, family, work, and an elderly mom in Africa, I only have enough time to watch amazing videos like these. Thanks so much Ken! Ina P.S. from another video I just listened to, I’ve had the imposter syndrome all my life 😁.
I think that is an awesome goal! Thank you for watching my videos, and I hope the future ones I make will be useful to you! Stories like these make me excited about producing more content!
That new intro of yours does look really cool, I have to say! :D I agree to the statement that DS has really changed over the past few years. Because it is such a young field I think that the landscape will change dramatically even for the next 5-10 years. My personal goal as a (now) first year university student is to stand out amongst my peers, more than 50% of every CS student seems to be interested in ML, but only a very small percentage of those really commit to the entire journey, which I really want to do. Have a nice weekend :)
Thanks for watching and great goal! I would definitely focus on some concrete steps to do that. For example: having a portfolio with 5 projects in it, or competing in 3 kaggle competitions, etc.
Thanks for the advice, I definitely agree with that! Currently I have about 3 ML-related projects in my portfolio and a few awards on my back which I mentioned on my personal website, so the next step for me at the moment is to expand my network and to get to know the people in this field a bit better! Kaggle is really great as well, I'll see if I can integrate that into my portfolio in the near future.
My Goal for the weekend : Understand existing 3 Power BI Reports. Learn 10 Excel functions and try it. Complete one micro course of Python on Kaggle over the weekend. Thank you it was a great piece of guidance.
Great video Ken! My goal: I want to pivot out of career in finance/advertising and into a career where I can answer questions in business or society via using numbers. My intellectual curiosity is not being scratched anymore in my current career. I want to dive headfirst into a field where not only is that intellectual curiosity being scratched but where I am also encouraged to explore, learn, and take risks.
Thanks for the video! You speak clearly! I worked in the Financial world for 10 years. Last year, I signed up for a course of Data Analyst and I loved it so much i decided to change my career and starting from zero. In my own experience, this thing of stablishing goals helped me a lot! I started with an online course of statistics, now i'm learning Python and starting a business project with a friend. Videos like this reenforces the steps we are taking and the path i choosed. Its great help! I will continue watching your videos! Thanks!
Love the new intro Ken !! Definitely agree with staring projects as soon as you can. I spent too much time learning about functions and things I could do but constantly forgot them because I was never doing projects and applying them. Alot of my time was spent learning things but I never knew when or how to apply them because I never did projects at first!
I 120% agree with the part "you go practice first and see if theory fits it". This is exactly how I would advise to other jr. data scientists that I train.
The problem here would just be if you don't go back to the theory. I mean this applies to multiple fields, but especially in data science it's important to understand why a Model behaves in a certain way. So try out everything, but make that by the end of it all you also understand what you are doing
@@frederikwitte9406 Completely agree here as well. That is why I recommend coding the algorithms from scratch. If you do this it is my belief that you have to understand the theory behind them.
Thank you so much bro, it’s very informative especially for someone who doesn’t any background on Data Analyst/ Science. Most likely I m receiving a lot of negative feedback that it’s too late for me to learn programming since I m too old. Gosh I’m turning 30 in few more months, time it’s ticking need to start soon. Wish me luck, hopefully I could start my journey and will work out finely then soon become successful in entirety.
I honestly don't think it is ever too late to start. There are plenty of people that learn how to code even into their 40's. I would focus more on the process of learning something new than worrying about the time it takes. I hope this helps!
Hello Ken, I am a doctor involved in clinical research and i stumbled upon your channel while trying to understand how to handle large clinical datasets. You are doing a great job of guiding us through data science
Sweet, right on time. I'm a Mech. Engineer just got my first job as an engineer, but I've lost total interest for it. Data science might be the career for me. Thanks for the tips. Good luck to everyone thinking of a career change or those studying! We got this!
@@ayushagrawal9633 Good actually! I'm not self-teaching, but was lucky enough to get into a masters program. I have an interview for a Jr analytics role that will hopefully get me one step closer to a data science position.
Good starting point. Let me share my experience. I have a master in statistics, I can code in R, and I have good experience to conceptualize data products. However, I'm still far from a data scientist. There's more about data engineering (getting the data ready) and deployment (platforms, UIs, APIs, servers, containerized, DevOps) that need to be included in my skillset. Your journey shows almost half the way towards data science. This is for everyone: keep learning new tricks and practicing with real examples.
I didn’t know about Kaggle micro courses. That will come in handy :) I have found that many people on Kaggle could have tough me a lot if they had documented their notebooks better. I learn a great amount by doing as much as possible from scratch at least once. Thanks for another good video :)
Thanks for watching! Yep, sometimes docs on kaggle can be rough. What I do is find people who document their code well and look at most of their workbooks. This saves me time on browsing!
Thanks for the tips Ken, especially on studying 30 min to 3 hours blocks. I started using Data Camp to learn about data science and have no idea what to do for beginner projects. Then I saw you had a video on this for kaggle. Definitely am going to check it out.
A lot of content over the internet about DS was uploaded more than 2 years ago. So sometimes I think that I lost my opportunity to begin DS career. So the thing that you upload your videos right now is the new light to me. Thank you
Finally RUclips algorithm got me to the right channel. I found the videos to be informative and thanks for sharing all these valuable links man! Kuddos!
I apply the concept of exciting accountability in my life, it involves making yourself accountable by telling people/the internet that youre embarking in a project, hopefully with a timeline, this pushes me to keep through. I did this by telling my friends and family I was learning how to code and that I was starting a YT channel and it has worked so far. I love learning skills so I will definitely check your Ultralearning video!
I've been learning data science for 3 months and I've been clearing my stats and r concepts and learning ml algos and after than I wanna jump into dl, nlp , ai though I'm from commerce background but I personally like it 😊
Great advice, and I think this video is great because it talks about the many different paths you could end up taking on a road into DS. The tip to do some real-world projects is key for understanding.
Oh wow !!! thanks too much for this video it makes me loving more data science. My Gol actually is in one year being able to work as a data scientist with python. I'm an financial (auditor) but python has been always a program that interest me too much specially the data science part. So please i'm open to learn and improve my self and the most have fun in this environment of learning.Again thanks you very much
@@parvjain3419 having learnt both, i feel Python is easier to learn than R.. but R feels more powerful.. not to mention u can use Python on Google's Collab and Notebook_ai when online.. saves you time maintaining your system.. just my opinion
@@parvjain3419 Both are just tools. Each have their own advantages. Personally I find both easy to learn. But python is very common. Just keep in the back of your mind to be program language agnostic..
Nice video! I'm not familiar with your channel but saw the thumbnail/its length and clicked through - a few other ideas/resources to choose from that I think can be really helpful for beginners depending on how they like to learn: --------------- 1. Make a habit of partitioning off a small portion of your learning time to do some light reading on best practices - this is softer information and can often be more subjective than the harder skills, but it can be extremely helpful in adding context to why certain things are the way they are and potentially help save you some cleanup down the road (things like docstrings, virtual environments, project directory structure, etc). (Like everything else) this can be overwhelming and my advice would also be to choose one thing at a time and start trying to build it into your workflow - docstrings and virtual environments are both good examples of things that take a little getting used to but can feel very fluid after a short period of time and save a lot of headaches with package managers (+ if you're using vscode or PyCharm the autodocstrings are great and keep your time spent doing the things the machine can't, not playing with margins) --------------- 2. Listening to podcasts about the subject (IMO) is also extremely valuable and can be done on the go - even if as a beginner you don't understand much that they're talking about, hearing others talk about common problems before you get to them can help you understand what's happening more quickly whenever you do. Additionally, a huge part of this field is learning how to articulate the problem you're trying to solve for on google/stack, and listening to podcasts can help you learn how to articulate things better both verbally and in writing. I'm much less familiar with the R podcast landscape but for Python beginners, Talk Python To Me is great for slightly deeper dives into a specific problem domain and Python Bytes is good/fun weekly exposure to the evolving packaging landscape. 2.1 Quick addition to the above point, another thing it can really help with is understanding the extent to which most problems, particularly on the data processing/cleansing/wrangling front have already been solved for - I've often seen more novice programmers try to recreate the wheel at every turn and hearing more experienced people unpack how they approach problem solving can help save you from writing a script to do something that's part of a standard library. --------------- 3. Time spent finding the right resources is never time wasted - in addition to paid courses/etc, I'd encourage finding someone on RUclips/similar who does a good job of articulating things in a way that makes sense to you. Corey Schafer has a fantastic Python channel covering everything from basics/best practices to end-to-end application development --------------- 4. Acknowledging that this would fall somewhere towards the latter half of the progression outlined in your video, I think it's super worth calling out that all the services we use personally have open APIs from which we can programmatically extract our data from and can be a great way to build out your mechanics in a fun/more personal way - it's also a natural and safer step towards web scraping since it's all the same communication protocol. Amongst many others, Spotify has a super user-friendly API and a well-supported helper library called Spotipy for anyone who might like this idea and be looking for a place to start Python beginnings to start learning how APIs work. --------------- Again great video and I hope the data dump's welcomed:)
It’s great to be able to watch and listen to someone who’s gone through what you’re currently trying to get through! I’m also going to do a 2nd masters (in Business Analytics) starting in the Fall and I’ve been trying to get a lot of practical experience via personal projects and some really good online resources from Coursera and Dataquest. It’s pretty cool to get reminded of the big picture like this and I look forward to watching the rest of your videos and new releases! Thanks for the awesome content and for providing clarity to those in pursuit of their dreams!
Can I ask why you’re getting a master in business analysis? Are a business analyst now? If not, what is your job and how long have you been in the analytics career path?
I have a history B's and I am basically starting from scratch self teaching myself about programing and data science. Thank you for providing multiple pathways to the end result because I cannot afford to return to college but have time and am willing to learn on my own time. I'm a lost 24 yo and seeing this calmly explained with lots helpful learning tips. This is what RUclips is still great for. Thank you....year old video so I doubt that you or anyone will see it but still thank you
Firstly, thanks for sharing your knowledge ! I live n Brazil and I'm graduate in Computer Engineer, recently signed a premium plan on DataCamp and I'm enjoying it.
I am venturing into Data Science field, hope this Video will help as it is my initial phase. Earlier I was working in travel industry in Product & Operations dept. (for 6-7 years). And pandemic proved need to do something else or more constructive by changing field to Technical.
This is such a cool guide. I'm studying data science and psychology at a university and this video is a great overview of what I can do over the summer or during down time in order to increase my knowledge and skill set. Looking forward to future videos!
Thanks for this video Ken. I almost gave up on data science but that's because I understood how bad my approach was after watching this video. Will try it your way and I'm sure I'll make progress. Thanks!
Within February 2021 I will learn concepts of probability and statistics and try to have a solid foundation on it and then I will jump into programming ( preferably R).
@@Dadum-bass Done with the probability part. Couldn't learn much on statistics though as I jumped into R programming instead. I completed 2 Udemy courses: Basic and Advanced on R. Hey! thanks for reminding 😀
@@sabindawadi741 thanks for being inspiring my friend. Seeing another make the shift into data sciences gave me confidence to switch from CNC machining myself
Hello, Ken. Very nice video, as usual. I've started my Data Science journey a few weeks ago, and I agree with your comments. I'm an engineer and work with computational mechanics for a while now, so I have a quite good understand about algorithms, programming, optimization, and stuff like this. What I figured out is that I already know some stuff about DS, just the terminology I was used to is different, which is great. I have in mind to create my own (much worst) version of scikit-learn, I mean, to implement the major ML algorithms myself, as you suggested. What I think it's also very useful is to create my own cheat sheets and reference notebooks, instead of just using some source already available, because this will force me to think more. Going through the packages' documentations and checking all functions that might be useful can also be a good idea, and this is slightly easier if the person has already some coding experience. I enjoy Coursera courses, and I'm relying on them to build my basic skills, but definitely what will help me most is to write my own codes. Cheers
Thank you for sharing your experiences, and recommendations if you did it all over. I just started my journey to Data Science, and have been studying from the ground up to prepare me for a pricey Data Science bootcamp this summer. I'll definitely be all over Kaggle, and try my own projects!
I think it will really help you to get ahead and learn some python like you had mentioned in your other comment. Most bootcamps are very intensive, and you want to do your best to prepare
I think learning by doing is a wonderful, fun and efficient way to learn a range of skills. HOWEVER, after 40+ years developing for large, medium and small commercial, non-profit, military/defense and other companies across communications, strategic data, media, AI, quantum-AI, full stack and other areas, I have also seen the dangers of that as a primary, let alone singular learning approach. People who learn exclusive via doing- especially working predominately on their own while eschewing theory and complementary skills prior to that approach, often have major applied theory and application context gaps in their knowledge that they appear generally unaware of. Obviously, this depends whether you are learning predominately on your own, or if you have guide(s), mentor(s) and/or teacher(s) that help point out short-comings and gaps in your knowledge base and approaches. There are multiple factors underlying this problem. First, it is very hard for most people to deeply self-introspect to see where they lie on broad based external knowledge continuums without significant sources of historical, theoretical and contextual reference. This is why the first factor of where people are on the Dunning-Kruger curve skews left with people who have largely self-educated without a broader base of external references as noted above. The second factor is the fact that most of the time, most people simply do not know what they don’t know. In other words, it takes a broad base of external theoretical, historical and contextual learning to develop metacognitive skills to help us intuit where we have gaps in our knowledge base, and to consistently discern correct from incorrect conclusions. While working on our own is a great way to learn deep, practical skills, we see that it often comes along with blinders to broader alternative views and approaches. In other words, you can easily be unwittingly blind to the fact that your toolbox is mostly bare, and that you don’t know how to optimally use even the tools you do have. For example, we see people who have never heard of Donald E. Knuth, JADD or Charles Petzold that have instead learned simply by doing, who code using approaches that others have long since bypassed, resulting in programming that causes problems and often must be rewritten and projects that must be taken over by other groups- ideally prior to them causing real world problems in the end user community. So, while there is no substitute for hands on learning and many have succeeded in becoming gainfully employed as programmers without much classical computer programming training or academic education; even hands-on learning approaches work better in the form of a balanced apprenticeship approach, where you work autonomously, but under the direction and oversight of someone much more knowledgeable and skilled than yourself. The fact is that the number of software developers on the planet doubles every 5 years. As such, by definition this means that half of software developers have less than 5 years experience. The number goes down as you get to more specialized areas of knowledge that increasingly can have a broad based impact in unforeseen ways on the future of humanity. For example, the number of AI programmers is doubling every 12-14 months and is still falling. This means that on average, AI programmers have a year or less programming skills in AI. Increasingly, the percentage that have any significant general programming or computer science education is dropping, as the demand for AI programmers is far outstripping what our universities and training centers can provide. As such, more people are self-educating, often by jumping in and doing prior to getting any significant training. Sadly, many of us who don’t get the academic knowledge and theory up front, never go back later to address that, since they get too busy and immerse in immediate needs. So, the gaps and deficiencies upon entering the workforce often are carried throughout their carrier. Often they become highly proficient programmers and such, in whatever language(s) they happen to need for the tasks at hand, but they often don’t learn to intuit when, for example they should turn to something like Node.js versus Python, or use Matlab vs Tensorflow, etc, So, what is the negative result of people learning predominately hands-on rather than under the oversight of others who are significantly more knowledgeable and skilled, and preferably with a good foundation of theory and broad based learning to bolster their metacognitive, theory and analytical skills development? The result is that suboptimal approaches, zero hour bugs, and higher rates of failed projects, as defined by the failure to meet one or more primary deliverables as defined by the final SOW. Further, as AI is integrating to the very fabric of our digital world, we are seeing a wide range of badly trained and architected systems that are leading to avoidable deaths, flash crashes in major global markets, just in time delivery system failures and much more. We know the causes of these are often programmers who are being primarily trained autonomous hands-on approaches. So, given that traditional education in the form of BS, MS and PhD’s in Computer Science and associated fields are increasingly becoming further out of reach of more and more people from a financial standpoint, what are the alternatives? First, online learning when done in conjunction with apprenticeship learning is a basis for a wonderful high utility educational base. Further, programmers need to step outside of their comfort zones and explore and identify areas that they are weak in. Also, as AI is integrating to the very fabric of our digital world, we are seeing a wide range of badly trained and architected systems that are leading to avoidable deaths, flash crashes in major global markets, just in time delivery system failures and much more. We know the causes of these are often programmers who are being primarily trained autonomous hands-on approaches. So, given that traditional education in the form of BS, MS and PhD’s in Computer Science and associated fields are increasingly becoming further out of reach of more and more people, what are the alternatives. First, online learning when done in conjuction with apprenticeship learning is a basis for a wonderful way to learn. Further, programmers need to step outside of their comfort zones and explore and identify areas that they are weak in. For example, say you spent a few hundred hours learning and taking on a few projects in Python and Swift. You should also read several of the seminal computer programming theory and related books- at the very least in the areas you are focusing on, and preferably add to them over time once you have internalized their benefit. Statistics repeatedly show that people who never learned core theory prior to beginning a career have a much lower chance of ever learning those areas. Similarly, it can be a great boon to identify a local universities or a company that needs interns where you can work under and along side others with different skills. While the pay may be minimal or non-existent, EVERYONE (self-taught, university skilled and all other modalities of learning) should intern- just as is done in other mission critical and related fields like medicine. And it is a LOT cheaper than the college alternative. In medicine it is done on top of 6-9 years of additional post-university graduation education, while in computer programming, a rapidly increasing percentage never even get any post graduation training of any significant merit- let alone intern/apprenticeship training. Is professional programming any less deep or broad of an area of knowledge? Not if you want to excel in terms of being at least well rounded and far more employable than average- and definitely not if you wish to excel in skills, compensation and job satisfaction. So, don’t migrate primarily to hands-on without getting a good base of other approaches behind you, because statistically you probably will never go back to rectify the competency, knowledge and related gaps you create by doing so.
Ken Jee I just started my masters in data science and I noticed that a lot of folks in this industry struggle to articulate the details. It’s not a mystical art, it’s science! Haha. Anyway, thanks again!
Explained here well!. I have 2 yrs as DA n last yr became DS. the amount of options you unlock by learning ml algos, stats, tools, visuals is endless. As a side hobby I'm developing predictive modelling on stock market data being a swing n long investor, pretty accurate estimates though. just need to include fundamental analysis n company prospects on top of it to make a call, earning good profits too. Its actually amazing we can do loads of stuff with Data Science, find out your passion n work on it!
Really awesome to hear that you made the transition! I would love your thoughts on my video "how to go from data analyst to data scientist", I hope the advice matches what you experienced!
@@KenJee_ds Hi Ken, I'm thinking of pursuing masters in DS in canada. what do you think is better, staying in my home country(India) or going abroad? I'm in pretty comfortable position here. Thinking of exploring the domain further in advanced tech and work few years abroad later.
Wow. Quarintine has made go for Data science from SAP which I am currently working on(Corporate job😁). I already started with DS Math skills to just revise what Maths I did during Engg. I also started with Python - everything from scratch was making it difficult. I watched one of your videos "Best free courses in DS". That has inspired me to go with what you have shared. Thank you for sharing your knowledge. Looking forward to more videos!!!
I'm trying so hard everyday , as day passes I feel I forgot what I have learnt ,I have already at it for a year and have a Post Graduate once I have completed ML and NLP but still not confident and cannot recall mainly NLP concepts , yet to do Deep Learning...really tough time , but I'm sure If I keep on at it I will crack it !! In three months my goal is to revise some 10 projects and deploy using flask ,testing in postmaster , exploring hiroku , NLP and Deep Learning will definitely down my life for sure !! But doing all this with a full job is even more horrible and keep drowning me ......
You can do it! You've done most of the hard work already! To be honest, I don't remember much of the concepts and stuff that I studied. When I start a new project, I essentially re-learn or refresh the material. I think this is pretty common for almost everyone in the domain. The best skill to learn is to pick things up quickly rather than to keep a vast store of knowledge. I hope this helps!
I'm studying Economics in College, probably my last year. I've been an Intern in Finance for ~3 years now (two different companies, also two different roles) and have just began feeling really interested in Data Science. If I could start over, I would pursue an internship in the field. I hope I can learn it and maybe even work with DS in the future. It is so much fun.
Subscribed! I really like how informative your channel is without being preachy. A lot of tech youtubers big themselves up to a point where you feel you're too late or too shit. Your channel provides useful resources and info. Thanks
I subscribed. Ken Jee, you know your stuff. Hopefully you don't lose your touch. As RUclipsrs become bigger, they loose their link to their viewers. As most social media apps does. Technology. To much data and to little heart!
This is the video I have been looking for! Thank you! I am going to come up with an algorithm to help me with my fantasy football draft this year. A tall order, I know, but I have a couple of months. We will see.
Great advice ken! Ill be going for bachelor of data science in the following month and this have been very useful as i wanna prepare before jumping into bachelors . Thanks for the video
Great video. I'm a mech engineer and have a couple of months in Codecademy data science and computer science. I have used Python before in a college projects of statistics and vehicles but now I really want to be able to feel condident saying I use it. Learning by doing projects and uppering the level is the best option, keep on with the good advice!
Thanks for sharing this. I've been kinda back and forth on my data science journey since I started a month ago. I started a trial with one learning site but had to quit because I don't have the funds now. Hopefully, I'll find a community here.
I highly recommend the free kaggle.com micro courses. I think that if you have enough motivation, you can learn all of the data science tools for free. Excited to have you as part of the community!
Thanks for watching everyone! Please leave some of your data science goals below so you can be held accountable by the community on this channel. Also, feel free to like and subscribe to help this channel grow! Watch this video on why I am starting data science over again next: ruclips.net/video/uXLnbdHMf8w/видео.html&ab_channel=KenJee ! Video on my #66DaysOfData Challenge: ruclips.net/video/qV_AlRwhI3I/видео.html&ab_channel=KenJee. Written version of this article: towardsdatascience.com/how-i-would-learn-data-science-if-i-had-to-start-over-f3bf0d27ca87
A few more links:
My Favorite Free Data Science Courses: ruclips.net/video/Ip50cXvpWY4/видео.html&ab_channel=KenJee
365 Data Science - Courses ( 57% Annual Discount): 365datascience.pxf.io/P0jbBY (Affiliate Link)
Bbvnb Van by
M mb.
.
porno
Hi Ken! Thanks for this. I wonder what you think about the 365 data science program. Do they consider real life data and projects?
01:20 - 1) Learn programming in python or R
01:40 - 2) Learn basic statistics for data science
01:58 - 3) Start projects as soon as you can (Kaggle)
03:28 - 4) Take notes on what algorithms and packages others use
03:46 - 5) Learn the source code for the algorithms and try to implement them from scratch
04:48 - 6) Work on more advanced projects where you collect your own data or use advanced concepts like deep learning, nlp, or computer vision.
Thanks for summarizing. I took this note and pasted on my wall 😃
Thank you
darling !!!
Thanks for the summary!
I don't know how to start a project. I've been reading, looking at others projects, but I'm still in the same place... I hate this feeling
Best advice from a data scientist: 1) nothing is what it seems. If you have yet to enter the data science and analyst field (still in school or studying), what you think a data scientist is day to day is wrong. So keep an open mind. We’re not writing algorithms on the windows of dorm rooms. Hint: 90% of our work is research and data preparation/cleaning. If you want to be a kick ass data scientist, learn how to interview data users. Coding is the easy part.
2) did I say be open minded? Good. Now tell yourself over and over and over, “almost no companies know what data science is”. And neither do you. So let me tell you what data science is: it’s the ability to understand and manipulate data and/or information. That’s it. Data science is not Python. If I give you an excel sheet of data and ask you to create a pivot that shows me counts and sums of those counts and you do that; you are being a data scientist. Your ability to code in a certain language does not = data science. I’m telling you this, cause being and looking this internet conceived data scientist part/person is a misconception and can kill your dream of ever growing into a great data scientist. Give yourself credit for what you know. And be confident that because you can analyze data in excel, you can also learn object oriented programming.
3) nothing is what it seems. Just cause a job title doesn’t say data scientist doesn’t mean your not going to do data science work. I’m a business intelligence developer. I manage data from Oracle, SQL Server and Cloud data lakes. When I sit down at my desk, I open Dbeaver, SQL Server, Visual Studio and Jupyter Notebook. All these IDE’s get used daily. That means I coding in 3 languages at a minimum daily. ONLY ONE of those languages is Python.
My job title is NOT data scientist.
My point in all my rubbish talk is that data science is a lot like the cloud. Over 10 years ago, the cloud concept became the rage. And yet most companies are still struggling to implement or even leverage the cloud. Data science is sadly going down that path too. Most companies and their managers are posting jobs to hire data scientists and as long as the job posting has a requirement of “must know Python”, it’s a data scientist job posting. However, 99% of the managers hiring data scientist don’t know what a data scientist is themselves. And those managers couldn’t tell you what a Jupyter notebook is. Or how to write/print “hello world” in Python. Yet they’re interviewing candidates for Python development. 😁
If and when you enter the so called data scientist career path, be open minded. This is a journey not a goal. Enjoy the trip. Be flexible. Knowledge trumps job titles. If you want to get paid the big bucks, be the most valuable player. That means work on everything sent your way. And show off your skills.
Good luck!
Great points! Being open minded and flexible are integral for success in this field!
Thank you so much, it's really helpful
Great comment. Very insightful.
this is a bit weird comment... you say your job is "90% of our work is research and data preparation/cleaning [...] That means I coding in 3 languages at a minimum daily. ONLY ONE of those languages is Python", and then "I’m a business intelligence developer [...] My job title is NOT data scientist".
I think you are indeed not a data scientist, and your job is not a data scientist job. This sounds a lot more like a data engineering job, or as you said, business intelligence developer. I'm sure it is related to data science, but I feel it's not quite the same.
@@dhidhi1000 I’m actually a Lead Data Scientist now. This is a bit of an old post. I think where we get lost in translation, is that I don’t work for one of the FANG companies. My company doesn’t blow money on a team of 5, all developing in one language out of VS Code. I have a team of 3. We support one department effort. And that means we not only own the data, we build the data, we run all analytics on the data. If the department we support needs to know the impact of a business decision, we perform the data science analyses, produce findings and reports, and build out the automation/bots. And not bragging, just being truthful; pure data scientist and certificate data scientist don’t last on our team. They usually tap out, because they are looking for a assembly line like work environment. They want to do one thing. For instance, they want me to “data engineer” and then hand them perfect data they can apply analytics to. We don’t do that. We are ninjas on my team. So 3 languages at minimum. SQL, Python, Java. And you also need to have a robust understanding of business operations. Spending time with the VP is not unusual. We make data scientist Managers and Director’s in short order. Because we bring our data scientists into all aspects of the business. I love our setting. Never a dull moment.
This guy replyin to each comment no cap. What a real G.
Doing my best! I get so many good questions, how could I not 😃
@@KenJee_ds sir I m MATH HONOURS GRADUATE should I do MSC IN DATASCIENCE.. I love coding but is it good for me
This is the first video I've been able to sit through engaged in a genuinely interested and joyful way on the topic of data science. You're not gimmicky or salesy. Thank you so much!
Thanks for the kind words! Glad you found it helpful!
I'm very young and I've been learning a lot of programming for many months. I had already started when I was 13, but nowadays I feel more enthusiastic than before. This kind of videos are really helpful and motivating. Thanks!
Glad to hear you are finding the videos motivating! It is awesome that you started so young!
I was very interested in Machine learning and data Science.
I would like to see all your videos. They may help me to achieve my goal. Please do many videos that encourage and guide the beginners.
Thanks 😊
I will! Thank you for watching!
Love this! As a pragmatist I find it appalling how intimidating this field appears before you start diving in and getting your hands dirty. Hopefully people can start to understand that DS is human intuition applying math and computing as a tool but not a concept separate from real word problem solving.
Definitely agree that it is just one advanced way of solving problems! Thanks for watching Matthew!
The thing I really like about your video is the stress on getting stuck into projects. I am a few months into learning about Machine Learning, but it is only since I started focusing on actually doing projects that I actually feel like I am beginning to understand what is going on. Before, I was focusing on programming, algebra, statistics, and I felt like I was just going through the motions, but now that I started trying to do the projects its like I am waking up and finally learning. Wish I had gotten stuck into projects much sooner.
Glad to hear that this method worked for you! I think it is definitely a paradigm shift, but after you do it you really reap the benefits!
Doing a complete career change from social services to data analyst with no experience or degree. This video has been very helpful. Thanks!
Awesome! Thank you for watching!
Hi, I'm about to start a masters in data science in September, this has been really useful!
Glad this has been useful to you! Thank you for watching!
What's your bachelor in
@@bodybuilding_updates i want to master in data science next year and im currently doing a BS in economics
am soo glad i found this when i was beginning my data science journey.. i am a physics major trying to venture into data science.. Thanks @Ken Jee
Glad you found it helpful! Thank you for watching!
I've learning data science for about 3 months now. My biggest problem was one those you mentioned in the beginner mistakes video : studying too much theory, trying out too many different courses . Your videos are great , man. Really helpful. Thanks. It made me more confident about making mistakes and trying out new projects . Not just kaggle , other more difficult platforms too.
This comment really made me happy. Thank you and I'm glad the videos have been helpful!
Bro plese reply me i am also think to start data science Mr farhan give me your ig so conect you and you help me
Always code on free time like the way you consume nonsense tiktok vids and other soc med material, you'll become a master in no time
So true!
Thanks Ken for a great video on the roadmap of learning data science! The intro animation graphics looks awesome!
Well said, accountability indeed helps to keep us motivated and pushing it to meet the goals we set. Being accountable for one another in this comment community is surely a great way to move forward in learning data science together. 😀
Thank you for the kind words! I am definitely loving the videos on your channel as well!
You man give some hope, because learning is really never ending and unless you hear it from an expert, it can be really discouraging. Thanks man.
Thanks for watching Tariq! It truly is a never ending process, but if you take it a day at a time I think things will work out!
Here's a course
Face Mask Detection Using Deep Learning . It's paid but it's worth it.
khadymschool.thinkific.com/courses/data-science-hands-on-covid-19-face-mask-detection-cnn-open-cv
I'm transitioning from a med student to data scientist. Currently applying to masters programs. Your videos have been really informative! Thanks!
Glad to hear they have been helpful! Thank you for watching them!
How is it going so far? I am a Mechanical Engineering student transitioning and I am looking to apply for a Masters programme.
Hi everyone. I am graduating from college this year with an economics degree. I have about a year of experience working as a data analyst (MySQL, Stata, R, Python proficient). Since startup recruiting has been a bit tricky because of the virus, I want to use the next few months while I am unemployed and living at home to become more proficient in higher ML techniques.
Thanks for watching and good luck!
My goal is to create a chatbot for my coaching business that will kick off on Udemy by next April. I just completed a month long Python boot camp and can now begin the real conversation design. My roadmap is in a paper I presented last September in Durban, SA - just as motivation to keep me at it. I’m a cybersecurity analyst and acquired my master’s degree in Linguistics at UGA in 2017. I guess I’m saying all this to force myself to stay on track and change careers to machine learning. No one around me (including hubby, kids or friends) is interested in machine learning enough to make this their second (non-paying) job, the way I have 👀-> as of now. Between home, family, work, and an elderly mom in Africa, I only have enough time to watch amazing videos like these.
Thanks so much Ken!
Ina
P.S. from another video I just listened to, I’ve had the imposter syndrome all my life 😁.
I think that is an awesome goal! Thank you for watching my videos, and I hope the future ones I make will be useful to you! Stories like these make me excited about producing more content!
That new intro of yours does look really cool, I have to say! :D
I agree to the statement that DS has really changed over the past few years.
Because it is such a young field I think that the landscape will change dramatically even for the next 5-10 years.
My personal goal as a (now) first year university student is to stand out amongst my peers,
more than 50% of every CS student seems to be interested in ML, but only a very small percentage of those really
commit to the entire journey, which I really want to do.
Have a nice weekend :)
Thanks for watching and great goal! I would definitely focus on some concrete steps to do that. For example: having a portfolio with 5 projects in it, or competing in 3 kaggle competitions, etc.
Thanks for the advice, I definitely agree with that! Currently I have about 3 ML-related projects in my portfolio and a few awards on my back which I mentioned on my personal website, so the next step for me at the moment is to expand my network and to get to know the people in this field a bit better!
Kaggle is really great as well, I'll see if I can integrate that into my portfolio in the near future.
Great! How are you doing now?
I liked how you outlined your goals! You crushed your subscriber goal for 2020! congrats!
I'm very grateful for all the support this year! Just shows what consistency on the platform can do!
My Goal for the weekend : Understand existing 3 Power BI Reports.
Learn 10 Excel functions and try it.
Complete one micro course of Python on Kaggle over the weekend.
Thank you it was a great piece of guidance.
Nice!
Came back to watch this during my study time. Working through the algo then coming back to the theory and fitting them together. Preach 🙌🏽
Also, just bought ultralearning. Here we go!!!
@@NicholasRenotte Let's gooo!!!
@@KenJee_ds I’m binging it right now. The fact that James Clear does the intro has got me sold already.
Great video Ken!
My goal: I want to pivot out of career in finance/advertising and into a career where I can answer questions in business or society via using numbers. My intellectual curiosity is not being scratched anymore in my current career. I want to dive headfirst into a field where not only is that intellectual curiosity being scratched but where I am also encouraged to explore, learn, and take risks.
Great goal!!
Would be interested to hear if you achieved this goal, and if so, how you are finding it?
Thanks for the video! You speak clearly!
I worked in the Financial world for 10 years. Last year, I signed up for a course of Data Analyst and I loved it so much i decided to change my career and starting from zero. In my own experience, this thing of stablishing goals helped me a lot! I started with an online course of statistics, now i'm learning Python and starting a business project with a friend. Videos like this reenforces the steps we are taking and the path i choosed. Its great help! I will continue watching your videos!
Thanks!
Thanks or the kind words and for watching the video!
Love the new intro Ken !!
Definitely agree with staring projects as soon as you can. I spent too much time learning about functions and things I could do but constantly forgot them because I was never doing projects and applying them. Alot of my time was spent learning things but I never knew when or how to apply them because I never did projects at first!
Thanks for watching! Very true about the projects!
Heyy brother it was such a relief to have found the deep wisdom you poured in through this video. I'm extremely grateful to you.
Thanks for the kind words about the video!
I 120% agree with the part "you go practice first and see if theory fits it". This is exactly how I would advise to other jr. data scientists that I train.
Glad you agree! I learned this later in my journey and definitely wish I had heard it from some where in the beginning.
The problem here would just be if you don't go back to the theory. I mean this applies to multiple fields, but especially in data science it's important to understand why a Model behaves in a certain way. So try out everything, but make that by the end of it all you also understand what you are doing
@@frederikwitte9406 Completely agree here as well. That is why I recommend coding the algorithms from scratch. If you do this it is my belief that you have to understand the theory behind them.
@@KenJee_ds exactly :-)
120%
That sounds like
Over fitting 🤨
Thank you so much bro, it’s very informative especially for someone who doesn’t any background on Data Analyst/ Science. Most likely I m receiving a lot of negative feedback that it’s too late for me to learn programming since I m too old. Gosh I’m turning 30 in few more months, time it’s ticking need to start soon. Wish me luck, hopefully I could start my journey and will work out finely then soon become successful in entirety.
I honestly don't think it is ever too late to start. There are plenty of people that learn how to code even into their 40's. I would focus more on the process of learning something new than worrying about the time it takes. I hope this helps!
You're literally describing what exactly I am doing to be a data scientist. Many thanks for confirming it.
This is what I love to hear! Thank you for watching Zahraa!
My goal this week is to finish a project on EDA for a FIFA dataset and learn about Ensemble Technique.
Great goal!
Hey great goal George, could you please share your study plan ?
i am learning Ensembling !!
@@anushrikadam5694 get in touch with me to know how
Congrats for 1M views! Many more to come, discovering your channel has been a blessing for me:)
Thanks Anubhav!
@@KenJee_ds 🍻
Hello Ken, I am a doctor involved in clinical research and i stumbled upon your channel while trying to understand how to handle large clinical datasets. You are doing a great job of guiding us through data science
Thank you!
Sweet, right on time. I'm a Mech. Engineer just got my first job as an engineer, but I've lost total interest for it. Data science might be the career for me. Thanks for the tips.
Good luck to everyone thinking of a career change or those studying! We got this!
I love the community support! Thanks for watching!
Good luck bro! Finance major here totally hating it so switching as well!
Great vid bro!
Hi @RamTheBeast! How you doing?
@@ayushagrawal9633 Good actually! I'm not self-teaching, but was lucky enough to get into a masters program. I have an interview for a Jr analytics role that will hopefully get me one step closer to a data science position.
Good starting point. Let me share my experience. I have a master in statistics, I can code in R, and I have good experience to conceptualize data products. However, I'm still far from a data scientist. There's more about data engineering (getting the data ready) and deployment (platforms, UIs, APIs, servers, containerized, DevOps) that need to be included in my skillset. Your journey shows almost half the way towards data science.
This is for everyone: keep learning new tricks and practicing with real examples.
Great story! You are 100% right, the learning never ends. I think that understanding is really important on the data science journey!
I didn’t know about Kaggle micro courses. That will come in handy :) I have found that many people on Kaggle could have tough me a lot if they had documented their notebooks better. I learn a great amount by doing as much as possible from scratch at least once. Thanks for another good video :)
Thanks for watching! Yep, sometimes docs on kaggle can be rough. What I do is find people who document their code well and look at most of their workbooks. This saves me time on browsing!
Thanks for the tips Ken, especially on studying 30 min to 3 hours blocks. I started using Data Camp to learn about data science and have no idea what to do for beginner projects. Then I saw you had a video on this for kaggle. Definitely am going to check it out.
Excellent! I also have a video on how to choose projects that may help! ruclips.net/video/yUrrf3Pm33s/видео.html&ab_channel=KenJee
Is datacamp a good platform?
Data science is a very vast field. Everyday I feel I know nothing and have a lots to learn. Thanks for the video. It is really helpful.
I couldn't agree more. Constant learning is the key. Thanks for watching!
@@KenJee_ds :) yes data science is definately a good way to curb an over active ego!
A lot of content over the internet about DS was uploaded more than 2 years ago. So sometimes I think that I lost my opportunity to begin DS career.
So the thing that you upload your videos right now is the new light to me.
Thank you
Glad to hear! Thank you for watching!!
Finally RUclips algorithm got me to the right channel. I found the videos to be informative and thanks for sharing all these valuable links man! Kuddos!
Glad the algorithm is working! Thank you for watching and for the kind words!
I apply the concept of exciting accountability in my life, it involves making yourself accountable by telling people/the internet that youre embarking in a project, hopefully with a timeline, this pushes me to keep through. I did this by telling my friends and family I was learning how to code and that I was starting a YT channel and it has worked so far. I love learning skills so I will definitely check your Ultralearning video!
Great stuff! I'll check it out!
My goal for this week: finish the SQL datacamp course, do one Cornell course of ML each day. Hope I can make it!
Great stuff! I believe in you!
After 20 years in IT mostly support, I am starting to learn data science and it’s my new goal, started learning sql, python and power BI.
Great stuff!
How many months you reckon it will take to become a data scientist if you do a full time study from University?
I've been learning data science for 3 months and I've been clearing my stats and r concepts and learning ml algos and after than I wanna jump into dl, nlp , ai though I'm from commerce background but I personally like it 😊
That is awesome!! Thanks for watching Rohan
you're doing it by your own rohan?
Great advice, and I think this video is great because it talks about the many different paths you could end up taking on a road into DS. The tip to do some real-world projects is key for understanding.
Thanks for watching Sean!
So cool to see someone with the same background as me (eerily so). Although you’re a few years ahead
Thanks for watching! Feel free to reach out if you have any questions. It seems like I can speak from experience haha
@Ken Jee how to reach out ?
What background is that?
Will software replace data scientists? Being able to do things with 3 clicks?
Oh wow !!! thanks too much for this video it makes me loving more data science.
My Gol actually is in one year being able to work as a data scientist with python.
I'm an financial (auditor) but python has been always a program that interest me too much specially the
data science part.
So please i'm open to learn and improve my self and the most have fun in this environment of learning.Again thanks you very much
That is an awesome goal, I think it is achievable with some hard work, good projects, and networking!
Thank you so much for recommending Kaggle. I had never heard of that site before. You may have just saved me a lot of money. lol
Glad it was helpful!!!
Awesome advice. I have just begun my DS journey and this gave a roadmap. Thank you Ken
Glad to hear this gave a solid roadmap!
my goal now is to learn DATA SCIENCE in the course of next 6 months
Great goal! Thanks for watching!
where do you plan on starting ? i am also in the same phase but just not know where to start
@@Peaceiscoming669 start with learning python or R
@@parvjain3419 having learnt both, i feel Python is easier to learn than R.. but R feels more powerful.. not to mention u can use Python on Google's Collab and Notebook_ai when online.. saves you time maintaining your system.. just my opinion
@@parvjain3419 Both are just tools. Each have their own advantages. Personally I find both easy to learn. But python is very common. Just keep in the back of your mind to be program language agnostic..
Nice video! I'm not familiar with your channel but saw the thumbnail/its length and clicked through - a few other ideas/resources to choose from that I think can be really helpful for beginners depending on how they like to learn:
---------------
1. Make a habit of partitioning off a small portion of your learning time to do some light reading on best practices - this is softer information and can often be more subjective than the harder skills, but it can be extremely helpful in adding context to why certain things are the way they are and potentially help save you some cleanup down the road (things like docstrings, virtual environments, project directory structure, etc). (Like everything else) this can be overwhelming and my advice would also be to choose one thing at a time and start trying to build it into your workflow - docstrings and virtual environments are both good examples of things that take a little getting used to but can feel very fluid after a short period of time and save a lot of headaches with package managers (+ if you're using vscode or PyCharm the autodocstrings are great and keep your time spent doing the things the machine can't, not playing with margins)
---------------
2. Listening to podcasts about the subject (IMO) is also extremely valuable and can be done on the go - even if as a beginner you don't understand much that they're talking about, hearing others talk about common problems before you get to them can help you understand what's happening more quickly whenever you do. Additionally, a huge part of this field is learning how to articulate the problem you're trying to solve for on google/stack, and listening to podcasts can help you learn how to articulate things better both verbally and in writing. I'm much less familiar with the R podcast landscape but for Python beginners, Talk Python To Me is great for slightly deeper dives into a specific problem domain and Python Bytes is good/fun weekly exposure to the evolving packaging landscape.
2.1 Quick addition to the above point, another thing it can really help with is understanding the extent to which most problems, particularly on the data processing/cleansing/wrangling front have already been solved for - I've often seen more novice programmers try to recreate the wheel at every turn and hearing more experienced people unpack how they approach problem solving can help save you from writing a script to do something that's part of a standard library.
---------------
3. Time spent finding the right resources is never time wasted - in addition to paid courses/etc, I'd encourage finding someone on RUclips/similar who does a good job of articulating things in a way that makes sense to you. Corey Schafer has a fantastic Python channel covering everything from basics/best practices to end-to-end application development
---------------
4. Acknowledging that this would fall somewhere towards the latter half of the progression outlined in your video, I think it's super worth calling out that all the services we use personally have open APIs from which we can programmatically extract our data from and can be a great way to build out your mechanics in a fun/more personal way - it's also a natural and safer step towards web scraping since it's all the same communication protocol. Amongst many others, Spotify has a super user-friendly API and a well-supported helper library called Spotipy for anyone who might like this idea and be looking for a place to start Python beginnings to start learning how APIs work.
---------------
Again great video and I hope the data dump's welcomed:)
The information is absolutely welcomed! Thank you for sharing!!
It’s great to be able to watch and listen to someone who’s gone through what you’re currently trying to get through! I’m also going to do a 2nd masters (in Business Analytics) starting in the Fall and I’ve been trying to get a lot of practical experience via personal projects and some really good online resources from Coursera and Dataquest. It’s pretty cool to get reminded of the big picture like this and I look forward to watching the rest of your videos and new releases! Thanks for the awesome content and for providing clarity to those in pursuit of their dreams!
Thanks for watching and for the kind words! I am glad that others are able to learn from my experience. Good luck on the rest of your degree!
Can I ask why you’re getting a master in business analysis? Are a business analyst now?
If not, what is your job and how long have you been in the analytics career path?
Just beginning to explore a career change into data science. Thanks for the great video. I'll be sure to watch some more!
Thanks for watching! I hope it was helpful!!
I can't say I'm good in math, but I can confidently say I'm not afraid of it. I'm excited.
That's a great attitude to have!
This is well said lol you got this
I have a history B's and I am basically starting from scratch self teaching myself about programing and data science. Thank you for providing multiple pathways to the end result because I cannot afford to return to college but have time and am willing to learn on my own time. I'm a lost 24 yo and seeing this calmly explained with lots helpful learning tips. This is what RUclips is still great for. Thank you....year old video so I doubt that you or anyone will see it but still thank you
Thanks for watching! Really glad to hear it helped!!
Watching the videos then reading the comments making me more excited to start.
Really happy to hear that! Good luck!
Firstly, thanks for sharing your knowledge ! I live n Brazil and I'm graduate in Computer Engineer, recently signed a premium plan on DataCamp and I'm enjoying it.
Thanks for watching! Great stuff, I would love to hear about what you learned from the course after you finish!
@@KenJee_ds Sure ! I will share my perspective with you after I finish the Data Science module. Keep the awesome job sharing your knowledge. Thanks
@@RicardoMeleiro Awesome!
Thanks for making this video Ken! It was really informative especially for a non STEM background student like me.
Thank you for watching Shagorika!
I am venturing into Data Science field, hope this Video will help as it is my initial phase. Earlier I was working in travel industry in Product & Operations dept. (for 6-7 years). And pandemic proved need to do something else or more constructive by changing field to Technical.
Glad to hear the video will help!
Your channel is like gold needle inside of haystack
Thank you for the kind words!
I completely agree
Ken Jee. Been Following you since you had 10 views on each video. I am SOOOOOO glad you are killing it. I already saw it coming
Thanks for supporting since the early days! I hope the videos are continuing to improve!
This is such a cool guide. I'm studying data science and psychology at a university and this video is a great overview of what I can do over the summer or during down time in order to increase my knowledge and skill set. Looking forward to future videos!
Glad the video was useful to you! Thank you for watching it!
Thanks for this video Ken. I almost gave up on data science but that's because I understood how bad my approach was after watching this video. Will try it your way and I'm sure I'll make progress. Thanks!
Really glad to hear you're improving on your approach! Excited to hear about your progress!
Within February 2021 I will learn concepts of probability and statistics and try to have a solid foundation on it and then I will jump into programming ( preferably R).
Awesome!!
Update?
@@Dadum-bass Done with the probability part. Couldn't learn much on statistics though as I jumped into R programming instead. I completed 2 Udemy courses: Basic and Advanced on R. Hey! thanks for reminding 😀
@@sabindawadi741 thanks for being inspiring my friend. Seeing another make the shift into data sciences gave me confidence to switch from CNC machining myself
Hello, Ken. Very nice video, as usual. I've started my Data Science journey a few weeks ago, and I agree with your comments. I'm an engineer and work with computational mechanics for a while now, so I have a quite good understand about algorithms, programming, optimization, and stuff like this. What I figured out is that I already know some stuff about DS, just the terminology I was used to is different, which is great. I have in mind to create my own (much worst) version of scikit-learn, I mean, to implement the major ML algorithms myself, as you suggested. What I think it's also very useful is to create my own cheat sheets and reference notebooks, instead of just using some source already available, because this will force me to think more. Going through the packages' documentations and checking all functions that might be useful can also be a good idea, and this is slightly easier if the person has already some coding experience. I enjoy Coursera courses, and I'm relying on them to build my basic skills, but definitely what will help me most is to write my own codes. Cheers
I think that is an awesome strategy! I would love to see the package you create once you get it up and running!
@@KenJee_ds It'll be my pleasure to share it with you :)
Thanks for the advice man, I'm leaving a like
I appreciate it! Thanks for watching!
Thank you for sharing your experiences, and recommendations if you did it all over. I just started my journey to Data Science, and have been studying from the ground up to prepare me for a pricey Data Science bootcamp this summer. I'll definitely be all over Kaggle, and try my own projects!
I think it will really help you to get ahead and learn some python like you had mentioned in your other comment. Most bootcamps are very intensive, and you want to do your best to prepare
@@KenJee_ds Yeah, that’s the plan. Thank you!
I think learning by doing is a wonderful, fun and efficient way to learn a range of skills. HOWEVER, after 40+ years developing for large, medium and small commercial, non-profit, military/defense and other companies across communications, strategic data, media, AI, quantum-AI, full stack and other areas, I have also seen the dangers of that as a primary, let alone singular learning approach.
People who learn exclusive via doing- especially working predominately on their own while eschewing theory and complementary skills prior to that approach, often have major applied theory and application context gaps in their knowledge that they appear generally unaware of. Obviously, this depends whether you are learning predominately on your own, or if you have guide(s), mentor(s) and/or teacher(s) that help point out short-comings and gaps in your knowledge base and approaches.
There are multiple factors underlying this problem. First, it is very hard for most people to deeply self-introspect to see where they lie on broad based external knowledge continuums without significant sources of historical, theoretical and contextual reference. This is why the first factor of where people are on the Dunning-Kruger curve skews left with people who have largely self-educated without a broader base of external references as noted above.
The second factor is the fact that most of the time, most people simply do not know what they don’t know. In other words, it takes a broad base of external theoretical, historical and contextual learning to develop metacognitive skills to help us intuit where we have gaps in our knowledge base, and to consistently discern correct from incorrect conclusions. While working on our own is a great way to learn deep, practical skills, we see that it often comes along with blinders to broader alternative views and approaches. In other words, you can easily be unwittingly blind to the fact that your toolbox is mostly bare, and that you don’t know how to optimally use even the tools you do have. For example, we see people who have never heard of Donald E. Knuth, JADD or Charles Petzold that have instead learned simply by doing, who code using approaches that others have long since bypassed, resulting in programming that causes problems and often must be rewritten and projects that must be taken over by other groups- ideally prior to them causing real world problems in the end user community.
So, while there is no substitute for hands on learning and many have succeeded in becoming gainfully employed as programmers without much classical computer programming training or academic education; even hands-on learning approaches work better in the form of a balanced apprenticeship approach, where you work autonomously, but under the direction and oversight of someone much more knowledgeable and skilled than yourself.
The fact is that the number of software developers on the planet doubles every 5 years. As such, by definition this means that half of software developers have less than 5 years experience. The number goes down as you get to more specialized areas of knowledge that increasingly can have a broad based impact in unforeseen ways on the future of humanity. For example, the number of AI programmers is doubling every 12-14 months and is still falling.
This means that on average, AI programmers have a year or less programming skills in AI. Increasingly, the percentage that have any significant general programming or computer science education is dropping, as the demand for AI programmers is far outstripping what our universities and training centers can provide. As such, more people are self-educating, often by jumping in and doing prior to getting any significant training.
Sadly, many of us who don’t get the academic knowledge and theory up front, never go back later to address that, since they get too busy and immerse in immediate needs. So, the gaps and deficiencies upon entering the workforce often are carried throughout their carrier. Often they become highly proficient programmers and such, in whatever language(s) they happen to need for the tasks at hand, but they often don’t learn to intuit when, for example they should turn to something like Node.js versus Python, or use Matlab vs Tensorflow, etc,
So, what is the negative result of people learning predominately hands-on rather than under the oversight of others who are significantly more knowledgeable and skilled, and preferably with a good foundation of theory and broad based learning to bolster their metacognitive, theory and analytical skills development? The result is that suboptimal approaches, zero hour bugs, and higher rates of failed projects, as defined by the failure to meet one or more primary deliverables as defined by the final SOW.
Further, as AI is integrating to the very fabric of our digital world, we are seeing a wide range of badly trained and architected systems that are leading to avoidable deaths, flash crashes in major global markets, just in time delivery system failures and much more. We know the causes of these are often programmers who are being primarily trained autonomous hands-on approaches.
So, given that traditional education in the form of BS, MS and PhD’s in Computer Science and associated fields are increasingly becoming further out of reach of more and more people from a financial standpoint, what are the alternatives? First, online learning when done in conjunction with apprenticeship learning is a basis for a wonderful high utility educational base. Further, programmers need to step outside of their comfort zones and explore and identify areas that they are weak in.
Also, as AI is integrating to the very fabric of our digital world, we are seeing a wide range of badly trained and architected systems that are leading to avoidable deaths, flash crashes in major global markets, just in time delivery system failures and much more. We know the causes of these are often programmers who are being primarily trained autonomous hands-on approaches.
So, given that traditional education in the form of BS, MS and PhD’s in Computer Science and associated fields are increasingly becoming further out of reach of more and more people, what are the alternatives. First, online learning when done in conjuction with apprenticeship learning is a basis for a wonderful way to learn. Further, programmers need to step outside of their comfort zones and explore and identify areas that they are weak in. For example, say you spent a few hundred hours learning and taking on a few projects in Python and Swift. You should also read several of the seminal computer programming theory and related books- at the very least in the areas you are focusing on, and preferably add to them over time once you have internalized their benefit. Statistics repeatedly show that people who never learned core theory prior to beginning a career have a much lower chance of ever learning those areas.
Similarly, it can be a great boon to identify a local universities or a company that needs interns where you can work under and along side others with different skills. While the pay may be minimal or non-existent, EVERYONE (self-taught, university skilled and all other modalities of learning) should intern- just as is done in other mission critical and related fields like medicine. And it is a LOT cheaper than the college alternative. In medicine it is done on top of 6-9 years of additional post-university graduation education, while in computer programming, a rapidly increasing percentage never even get any post graduation training of any significant merit- let alone intern/apprenticeship training.
Is professional programming any less deep or broad of an area of knowledge? Not if you want to excel in terms of being at least well rounded and far more employable than average- and definitely not if you wish to excel in skills, compensation and job satisfaction. So, don’t migrate primarily to hands-on without getting a good base of other approaches behind you, because statistically you probably will never go back to rectify the competency, knowledge and related gaps you create by doing so.
an interesting read, thank you for the insight
Thank you so much for coming up with this video! This is definitely impactful! My first video of this channel and hi from your subscriber!
Thanks for watching and for the sub! Hopefully I will have more impactful ones coming soon!
Love the way he describes getting into this field!
I appreciate it!
Ken Jee I just started my masters in data science and I noticed that a lot of folks in this industry struggle to articulate the details. It’s not a mystical art, it’s science! Haha. Anyway, thanks again!
Explained here well!. I have 2 yrs as DA n last yr became DS. the amount of options you unlock by learning ml algos, stats, tools, visuals is endless. As a side hobby I'm developing predictive modelling on stock market data being a swing n long investor, pretty accurate estimates though. just need to include fundamental analysis n company prospects on top of it to make a call, earning good profits too. Its actually amazing we can do loads of stuff with Data Science, find out your passion n work on it!
Really awesome to hear that you made the transition! I would love your thoughts on my video "how to go from data analyst to data scientist", I hope the advice matches what you experienced!
@@KenJee_ds Hi Ken, I'm thinking of pursuing masters in DS in canada. what do you think is better, staying in my home country(India) or going abroad? I'm in pretty comfortable position here. Thinking of exploring the domain further in advanced tech and work few years abroad later.
This is an amazing channel. Glad I found it !!. I’m planning on majoring in data science this fall, thank you for all the videos .
I'm glad you found it too! Thank you for watching!
Oh my god my school (a pretty renowned one) doesnt offer that!
Wow. Quarintine has made go for Data science from SAP which I am currently working on(Corporate job😁). I already started with DS Math skills to just revise what Maths I did during Engg. I also started with Python - everything from scratch was making it difficult. I watched one of your videos "Best free courses in DS". That has inspired me to go with what you have shared. Thank you for sharing your knowledge. Looking forward to more videos!!!
Thanks for watching! I am glad the videos have been useful to you!
Video starts from 1:15
thank u
lol i was about to post this
All I can say is thanks
Thanks bhai
Thank you for providing your insights its always good to have some guidance from the people already in the domain.
Thank you for watching!
I'm trying so hard everyday , as day passes I feel I forgot what I have learnt ,I have already at it for a year and have a Post Graduate once I have completed ML and NLP but still not confident and cannot recall mainly NLP concepts , yet to do Deep Learning...really tough time , but I'm sure If I keep on at it I will crack it !! In three months my goal is to revise some 10 projects and deploy using flask ,testing in postmaster , exploring hiroku , NLP and Deep Learning will definitely down my life for sure !! But doing all this with a full job is even more horrible and keep drowning me ......
You can do it! You've done most of the hard work already! To be honest, I don't remember much of the concepts and stuff that I studied. When I start a new project, I essentially re-learn or refresh the material. I think this is pretty common for almost everyone in the domain. The best skill to learn is to pick things up quickly rather than to keep a vast store of knowledge. I hope this helps!
Liking your new intro and B-roll!! Thank you so much for the advice.
Thanks! I think it was definitely worth it to up the production value!
I'm studying Economics in College, probably my last year. I've been an Intern in Finance for ~3 years now (two different companies, also two different roles) and have just began feeling really interested in Data Science. If I could start over, I would pursue an internship in the field. I hope I can learn it and maybe even work with DS in the future. It is so much fun.
You can definitely make the transition later from your career in finance if you wanted to! A lot of people I know have
I am a new joiner in data science and was browsing about it and stuff. Saw your video, click on it, subscribed, and shared with friends.
Thanks for the support! I hope you find the videos helpful!
@@KenJee_ds Yes bro, really it is. I am making notes also.
Kaggle - recommended website for datasets, learning data science
Agree!
Subscribed! I really like how informative your channel is without being preachy. A lot of tech youtubers big themselves up to a point where you feel you're too late or too shit. Your channel provides useful resources and info. Thanks
Thanks for watching! Honestly, I was late to the game and still made it into the field. I don't see why other people can't do it too!
My goal is to deploy a model by the end of this year
Great goal! Thanks for watching!
Great. Liked the clean chart about 'Plan of Action'. Good.
Thanks for watching Venkat!
He is one of those good youtubers that people subscribed after watching one video
Thanks!
Facts. I did the same after one vid
I subscribed. Ken Jee, you know your stuff. Hopefully you don't lose your touch. As RUclipsrs become bigger, they loose their link to their viewers. As most social media apps does. Technology. To much data and to little heart!
Great video Bud. I really appreciate it !!!
Thanks for watching Shakti!
Believe me, bro, you will have 1 m subscribers soon, keep going good luck
Thank you for the kind words! I am just taking it a step at a time and enjoy the journey!
This is the video I have been looking for! Thank you! I am going to come up with an algorithm to help me with my fantasy football draft this year. A tall order, I know, but I have a couple of months. We will see.
I love it! Can't wait to hear how it goes. If it works I will be asking for picks haha
@@KenJee_ds No problem at all. I will definitely be sharing my model and code in Kaggle!
I literally was searching for this kinda advice a few days back and so happy I found this video! Thanks for the advice and insights :)
Glad you found it! Thank you for watching!
Great advice ken! Ill be going for bachelor of data science in the following month and this have been very useful as i wanna prepare before jumping into bachelors . Thanks for the video
Thanks for watching!
Thanks ,I need to learn the data science and currently i'm pursuing b.tec and i'm good at python also very passionate about to learn data science
Awesome!
I like that this is not an overly technical video and drew me in to DS instead of being scary with loads of terminology
Thanks for watching! I do my best to try to explain things in plain english rather than using data science buzzwords
I wish there was a google classroom for beginners who are studying data science with a expert in the group who could guide us😓
I think there are a few programs like that online! Will try to post them when I see them
@@KenJee_ds tysm❤️
Great video. I'm a mech engineer and have a couple of months in Codecademy data science and computer science. I have used Python before in a college projects of statistics and vehicles but now I really want to be able to feel condident saying I use it. Learning by doing projects and uppering the level is the best option, keep on with the good advice!
Thanks for watching Jullien! Keep up the project work!
- Get familiar with R and then Python
- Learn in depth statistics
- Apply what I learned for a better understand of sports
Yep!
thanks friend
Thanks for watching!
I would like to develop my python and numpy skills in order to prepare for a masters in data science after my third year of university
Great stuff! Thank you for watching!
Thanks for sharing this. I've been kinda back and forth on my data science journey since I started a month ago. I started a trial with one learning site but had to quit because I don't have the funds now. Hopefully, I'll find a community here.
I highly recommend the free kaggle.com micro courses. I think that if you have enough motivation, you can learn all of the data science tools for free. Excited to have you as part of the community!
@@KenJee_ds Hi Ken, thanks for your response. I opened an account with Kaggle today and I'll start working on it from tomorrow. Thanks again 👍🏽
@@solaawodiya7360 Great stuff!
Congratulations for 1 Million views 👏🙌
Thanks Piyush!
Thanks, you really did cover all the topics that I've been looking for. Great work :)
Thanks for watching Nimish! Glad the video was useful!
I wish I had access to this sort of video when I was starting my career.
I wish I started making them earlier haha! Thanks for watching!