5 THINGS I WISH I KNEW Before Starting Data Science 

Thanks so much @linz linz! Honestly I went through it all myself not too long ago!

Thanks so much @wwwdenis, my core goal is to make it a lot more pragmatic! And ultimately allow people to learn and use useful tech easier! I come from a non-tech background so I try to explain it how I wish somebody had explained it to me when I was learning 😃.

Ohhhh man, thank you so much for your kind words @Grejegando!

Well said. I wish you success in any field you get into as well.

Hell yeah, would've saved a bunch of time at the start! Thanks @Sean, great vids on your channel btw! Loading 1m rows into SQL...hell yes!!

Woahhh 🙏🙏 Thanks so much @Rui, definitely I've got some more interactive content ideas planned!!

Heya @ismail, I’m glad you got some value out of it!! Let me know if you’d like to know anything else!

Anytime @Bayu!

Stay tuned for the siamese network vid @Lukas, I'll show how to integrate a hardcore model into a Kivy app.

Cheers @Crystal!

Thanks so much @tasarim, sharing a little knowledge is a huge reward in my books!! Everything else is a bonus 😊

Thanks so much🙏! Honestly, it's a great skill to have in your back pocket particularly if you're already well versed in SE or Development!

Thanks so much @Deepak, so glad you enjoyed it!

Hey Deepak, I am a fellow data scientist in Australia just starting out on RU-vid too - just posted my first "a day in the life" video and would love for you to check it out! Also will be posting other videos soon on what I do/ interview tips/ job market in Australia :)

Awesome stuff @Om!

Thanks a bunch @Omlettovore!

Exactly!

Thanks so much @@zaheersuhabuth2677 !

Hey Omelettovore :), I am a fellow data scientist in Australia just starting out on RU-vid too - just posted my first "a day in the life" video and would love for you to check it out! Also will be posting other videos soon on what I do/ interview tips/ job market in Australia :)

Thanks soo much! Glad you enjoyed it @Shivan, plenty more to come!

🤣 thanks so much @Stas!

Thanks so much @Jie!

Thanks so much @NHMI!

YESSS! Thanks so much @Rodrigo!

Hey Rodrigo, I am a fellow data scientist in Australia just starting out on RU-vid too - just posted my first "a day in the life" video and would love for you to check it out! Also will be posting other videos soon on what I do/ interview tips/ job market in Australia :)

Heya @Cheska, I mainly look at documentation for a particular thing I'm working on! I keep a big list of all of my ideas/projects I want to build then: 1. On Monday's I'll plan two projects for the week and start researching how to build them 2. Every day, I'll take a look at documentation or examples that are in a similar vein and build from there E.g. Atm, I'm doing a lot of work with Mediapipe so the documentation (google.github.io/mediapipe/) and the GitHub repo has been my go to (github.com/google/mediapipe).

YESSS, awesome work @Paulina! It is literally the best book ever, my friends are definitely over me talking about it all the time but meh 😅

Thanks so much @Amir!

I've heard they're very good, I haven't done any myself but have some of the stats ones planned as a refresher!

MOOCs all the way!! If you're a physicist, you'll be fine with the math just need to tighten up on the DS specifics.

Heya @John, you're in great stead if you have CS as your foundation knowledge. I found the data science majors more interesting when I started my masters but stats will still provide you with a great baseline. If I had to pick out of those two though I'd go with DS all day.

Heya @Somdutta, the fundamental skills of programming and working with data are unlikely to go out of fashion soon! Whilst there have been numerous 'hot roles', be it Business Intelligence Specialist, Big Data Engineer, Data Scientist or Deep Learning Engineer, the fundamental skills behind them have been some form of programming and an understanding of data. My guess is that these skills will continue to be in demand into the foreseeable future, the name of the role that they're attached to will likely change though. So...after all my rambling, my advice is to invest in hard skills. If I had to go back to university now I would invest in a Computer Science degree rather than a specialist degree like data science. You can always have a major in DS or Deep Learning from a Comp Sci degree. Having a degree varies greatly depending on where you're looking to work, take a look at some job ads in your region to see what they're looking for! Before committing to a degree, I'd highly recommend you go through my DS crash course playlist ru-vid.com/group/PLgNJO2hghbmjNrHZqplNMEpsW-QLFdvJV. This style of work is the stuff I repeatedly do with clients!

@@NicholasRenotte I am a civil engineer. Can I learn data analysis and find. a job or it must be relative to a bachelor's degree in computer to work in it ?

Heya @Yann, same languages as me! Given you already have some Comp Sci experience, I'd suggest building out your portfolio with more data science assets. Try your hand out at some machine learning projects, I'd suggest taking a look at my full stack data science tutorial, I go from raw data to modelling and finally building an integrated ML React App: ru-vid.com/group/PLgNJO2hghbmjNrHZqplNMEpsW-QLFdvJV Once you're comfortable, you could probably start by going into data analyst or junior data science roles and move into a senior role pretty quickly.

@@NicholasRenotte Thank you Nicholas ! I appreciate your help very much. I will watch your video very soon :)

@@YannSchmidt anytime @Yann, awesome!

Awesome!! Honestly, try some DS projects out and see what you think of it. Bunch of fun tutorials to try out on the channel, some of my beginner friendly favs: 1. Build a Grammar Correction App using Deep Learning: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-4htP-hmIBvc.html 2. Bicep Curl Counting using PoseEstimation: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-06TE_U21FK4.html 3. Landing a Space Shuttle using Reinforcement Learning: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-nRHjymV2PX8.html

@@NicholasRenotte Thank you so much! I appreciate it!

Awesome! So glad you enjoyed it @Fred!

It's a mix, a large number of orgs now use open source in collaboration with DnD tools like SPSS and AutoML!

Heya @luiza, I think the same applies to many jobs given the current environment. My hours tend to fluctuate, during busy periods they tend to shoot up but then I also have a reasonable amount of down time where I can skill up and take it a little bit easier. I still get time to do the things I enjoy almost every day, run, swim and go out to dinner. Have you checked this out: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-bVCrzoIVudU.html

Hey Luiza, I am a fellow data scientist in Australia just starting out on RU-vid too - just posted my first "a day in the life" video and would love for you to check it out! Also will be posting other videos soon on what I do/ interview tips/ job market in Australia :)

Thanks 🙏🙏🙏 @Pratik, we're getting there though 😊!

Heya @seon, rather than taking a huge leap like college, start taking some free online courses or check out some tutorials! Get comfortable with data, learn Python and start building stuff! This playlist walks you through end to end how to work with data and build ML models, let me know how you go: ru-vid.com/group/PLgNJO2hghbmjNrHZqplNMEpsW-QLFdvJV

Awesome stuff @Eugene! Keep it up, you'll be fine with the math, just take it bit by bit!

@@NicholasRenotte I actually started relearning everything at my own pace. Starting from rational expressions. I was surprised of the result. Removing the time and pressure factor of traditional university setting, and just patience and youtube and google lessons, I was able to solve polynomials that would have taken me days or weeks if I had solved it 10 years ago.

@@eugenel8805 agreed, I think taking the pressure away and having a beginner mindset makes all the difference!

Hmmm interesting question, I personally think you should have an interest in it when you begin. That's not to say you can't grow to love it. Actually now that I think about it, I kinda had to force myself to learn at the beginning, I did like programming but I sucked at it. Do you have an interest in data science? If so, at least that's a start. Try your hand out at some coding and see how you go and if you like it!

I actually had a think about this a few nights ago, there really isn't anything which is an industry standard. I went through a lot of the deeplearning.ai courses on Coursera when I was starting out but it wasn't a mandatory requirement for my job but they are very good from a theory standpoint! Right now, I'd say take a look at the Tensorflow Developer Certificate and possibly some of the stuff from OpenCV as well! Yep, I tend to use both MacOS and Windows. Honestly, my advice is use what you're comfortable with but have an understanding for how to use both as different clients tend to use different stacks and architectures.

@@NicholasRenotte thank you, very insightful

@@hareshrk6259 you're welcome, let me know if you end up doing any of them!

@@NicholasRenotte haha sure I'm still doing my basic python rn, still have quite some time before I do any certificates

@@hareshrk6259 all good, baby steps, I'm excited for you to see all the AMAZING Python stuff there is out there!

Heya @Gary, sure am! I started my Masters but ended up putting it on (indefinite) hold.

😂🦌bambi eyes checking in!

🤣

Definitely, I'll add it to the list. It's mainly about skills @Abdullah, if you can show to your peers that you can do the work then honestly that's all that matters. Certificates are great for proving that but practical experience and a portfolio are just as good if not better!

Hit me up on Discord Emmett, always happy to chat!

@@NicholasRenotte Thanks man! :D Is there a link to your discord on your channel? I couldn't find one

@@emmettmulroy6138 bit.ly/3dQiZsV :)

Great choice @Jaskeerat, honestly they're so powerful when used together!

Hmmm, what are you interested in? This might be useful as a start: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-oLpBGtY-_sI.html

@@NicholasRenotte thanks for the reply, turns out I have watched this vid. very useful.

@@nontombisiro149 awesome stuff. Well in that case, I'd extend our your GIS skillset. Perhaps start looking at geospatial ML and mathematical optimisation for route based problems!

@@NicholasRenotte thank you

@@nontombisiro149 anytime!

Give me a few weeks 😉

Baby steps @Sajil, check this out for where and how to start: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-oLpBGtY-_sI.html

Honestly, I have no idea. Though, given what I'm seeing in the industry, where current investment is going and the languages major cloud platform providers are supporting, I would say no. Most that I work with only support R, Python, Scala. And the large majority of innovation in the space is being done in those same languages with the addtion of C++. I think you're probably safe with Python for now but as always, the space is ever evolving so a little bit of continuous learning each day future proofs your career so it doesn't hurt to learn Julia if you're interested.

@@NicholasRenotte Thank you for your insight. I agree with you that Python is a safe bet for the time being, even if it goes down, it won't drop suddenly, probably takes many years to sunset. As you rightly said, C++ is been with us since 70s, it hasn't completely disappeared despite new languages popping up.

@@ButchCassidyAndSundanceKid exactly, in fact I've still got colleagues who are now even more in demand because they have COBOL skills. You're right, there really isn't a cliff in demand, particularly for those languages that have been around for a long time!

@@NicholasRenotte One of my ex-colleagues is still programming IBM midrange AS400 in 2021 !!!

@@ButchCassidyAndSundanceKid damn 😂

Nope! I think it's a useful tool to help network but if you don't like it try growing your profile on GitHub (particularly if you're going for a developer role).

Definitely, you could specialise in DS from within your CS degree as well!

Ah it's normally a bit of a mix, for standard tasks I'll use benchmark datasets e.g. KITTI, MNIST, ImDB (en.wikipedia.org/wiki/List_of_datasets_for_machine-learning_research). For custom tasks I'll either collect the images myself using a camera, drone or webcam/OpenCV or I'll scrape data using Beautiful Soup. In terms of tools were you referring to like coding tools/libraries or hardware style stuff?

@@NicholasRenotte In terms of tools I mean like rapid miner or orange for data mining .So you use the application Beautiful soup on Python to get your data am I correct?

@@baryonyx0387 Oh got it! So, this is my stack below based on the task: Deep Learning - Tensorflow, Pytorch, Keras Traditional ML - Scikit-Learn, Pandas, Numpy, Seaborn and Matplotlib, Scraping Data - Beautiful Soup Drag and Drop Modelling - SPSS Modeler, Watson Object Detection Computer Vision - OpenCV, LabelImg, Tensorflow Object Detection API NLP - Watson Natural Language Understanding, Watson Assistant, Watson Discovery Application Development - I mainly code in Javascript for Web Dev with React, Redux, Tensorflow.JS, (starting with ml5.js soon) Reinforcement Learning - OpenAI Gym, Keras-RL Future stuff I've got planned on my list mainly revolves around learning C++ and Unity for advanced reinforcement learning use cases!

@@NicholasRenotte big help fam thanks

@@baryonyx0387 anytime my man!!

Heya @Amir, there's math in pretty much all of it, but some of it is relatively basic whilst other components can be more difficult. With hardcore deep learning you're likely to need more math than most. If you're working on viz or data exploration it will be less advanced.

Of course. Having a solid grasp of statistics is going to make your life very easy. More at basic level, being conversant with linear algebra and calculus is a must if you're serious about this.

Oh, but it makes it soo much easier to learn other programming languages later on @Fellipe! I'm actually beginning to learn C++ now!

@@NicholasRenotte hehe I'll give it a try! Thanks bro!

@@fellipejacob87 awesome stuff, let me know how you go!

Freetalk! 😊

Heya @favour, check this out: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-oLpBGtY-_sI.html

@@NicholasRenotte is it possible for me to start it with my phone because i don't have pc and am using pydorid on my phone as editor for python

@@favourwilliams3771 it might be a little tricky as you won't have access to a lot of the functions from a regular machine but give it a go!

@@NicholasRenotte ok so me learning python now is good to go? like to start with that?

@@favourwilliams3771 definitely, I'd start with getting really good at Python!

I think you have to really enjoy it, nothing ever worth doing was easy! You just have to take that first step ✌🏽

Awesome having insight from the recruiting side. Do you find that a masters is needed because of corporate policy or because the calibre of individuals is commensurate with their formal qualifications? I know the US highly values tertiary education so this is very much in line with what I'd expect but I'm always interested in hearing why.