Экстракт Знаний

Do you have what it takes to be a great data scientist?

8:28

Do you have what it takes to be a great data scientist?

Tina Huang 20.12.2020 9 612 просмотров 458 лайков обн. 18.02.2026

Смотреть на YouTube

Поделиться Telegram VK Бот

Транскрипт Скачать .md

Анализ с AI

Описание видео

As an entry level FAANG data scientist with a few months under my belt, I've observed many data scientists and even some great data scientists. In this video, I go through the 5 traits of a great data scientist. If you're deciding if data science is right for you or want to know if you have what it takes to be a great data scientist - this video is for you! Or if you're already a data scientist, keep watching to see how you can level up! ______________________________________________________________________ You might also be interested in these videos: Day in the life of a FAANG Data Scientist: https://www.youtube.com/watch?v=lCi6fWuI8r4 How I learned SQL from Scratch in 11 Days to Pass my FANG SQL Interview: https://www.youtube.com/watch?v=vaD3ZFFNwhM ______________________________________________________________________ Subscribe: https://www.youtube.com/channel/UC2UXDak6o7rBm23k3Vv5dww/?sub_confirmation=1 ______________________________________________________________________ Check out StrataScratch for SQL interview prep: https://stratascratch.com/?via=tina ______________________________________________________________________ Contact: youtube: youtube comments are by far the best way to get a response from me! I answer every single comment! linkedin: https://www.linkedin.com/in/tinaw-h/ (second preferred but I might suck at responding) email: hellotinah@gmail.com *If you're reaching out through linkedin or email, please leave a youtube comment just letting me know that you reached out :) ______________________________________________________________________ *The StrataScratch affiliate program give me a small portion of the sales price at no cost to you. I'm currently not monetized and really appreciate your support in helping improve this channel! :) #DataScience #TinaHuang

Оглавление (7 сегментов)

0:00 Intro 216 сл.
1:08 growth mindset 300 сл.
2:44 unbiased curiosity 283 сл.
4:16 paranoia 260 сл.
5:28 organization 128 сл.
6:11 storytelling 221 сл.
7:16 BONUS: business domain 266 сл.

Intro

now that i'm a few months into my data science job i'm lucky to have interacted with many data scientists across a bunch of different teams and i've also had the honor to interact with amazing data scientists that i really look up to in this video i wanted to share with you guys my observations on what makes a great data scientist this is not just what the minimum skills are to get into data science like math or programming if you guys are interested in that i can also make another video about it this video is for those of you that have high aspirations for yourself when deciding if data science is the right career choice you don't want to be just a mediocre data scientist or if you're already a data scientist what you can do to level up so now without further ado these are the five things that make a great data scientist in my opinion number one is growth mindset number two unbiased curiosity number three paranoia about the data number four organization and number five storytelling and a bonus one which is choosing the right domain now in my opinion the most important trait that differentiates a great data scientist from an average one is a growth mindset

growth mindset

is when you believe that you're able to improve and learn you might have to work really hard on it but you know that you get there now contrast this with a fixed mindset where you believe that your skills and abilities are predetermined and set in stone and this is just the way you are forever now a growth mindset is incredibly important because the speed at which a data science field is evolving is so incredibly fast new data science technologies are literally being built faster than you can possibly learn them and the standards of how to analyze data itself is shifting every day as well when i was working in bioinformatics one day everybody was you know happily working in art and then it felt like suddenly it was all about python because the ml packages are just they're just so much more powerful and easier to use so guess what i had to go and learn python and the corresponding packages now if i had a fixed mindset it would be something like oh no i don't really know python i don't understand this documentation and the syntax is so confusing i must be too stupid i should just give up now these days with new things popping up left right and center like automl gpt3 etc you get where i'm coming from kind of hard if you have a fixed mindset what i'm like with my growth mindset though is something more like this oh no i don't really know python i don't understand this documentation and the syntax is so confusing i must be too stupid so i need to get less stupid yay process syndrome is real my friends but that's a topic for another time unbiased curiosity you ever know some

unbiased curiosity

people that have really strong opinions about things without you know really thinking about it like for somebody that just truly believes that tesla stock would skyrocket do a stock split and then continue to skyrocket again which totally did happen or someone that just truly believes that coconut oil is a miracle drug it can cure cancer and all diseases well if you're either of these people then data science is probably not for you let's deconstruct unbiased curiosity unbiased is when you go into something with an open mindset and craft the story from the data yes you should always have hypotheses to test out but it's a big no-no to go in with already formed opinion and just try to gather data to support your opinion hey your opinion can be very well amazing and correct like our friend here that invested all their money in tesla became overnight millionaire but that wouldn't make you very good data scientists a data scientist needs to be unbiased and let the data shape their opinions now the second component is curiosity you have to be a curious person the best data scientists that i've met are the ones that always ask themselves questions why is this metric always falling what other products do our customers use why is mitochondria the powerhouse itself no that's just me that led me to four years of undergrad degree in science to answer that question well when you get unbiased plus curiosity you know what that makes a hella good scientist and what is the second part of this job title oh it's scientists judge i rest my case paranoia this was the one that was

paranoia

really surprising to me when i was not creepily observing great data scientists at work i noticed a common trait among them is paranoia not that they think that the government is out to get them i think not at least have you guys heard about sanity checks it's when you write some code with some constraints and queries to make sure you haven't gone and messed something up the problem with data science is that even if you do something wrong you'll still get results provided that you didn't make a syntax error an example that i'm very embarrassed to say i have done is something like basically messing up the denominator or fraction i was trying to compute i did like sum instead of count and sql to be exact and well if i did a sanity tag i would have then realized that the number of distinct rows was completely off but i didn't and this was extra embarrassing because a senior data scientist caught this for me in fact though i was lucky that the data scientist even caught it for me it can be very bad if you're not constantly checking yourself because some things are really hard to catch and realistically nobody's gonna go through your code line by line to catch it for you and then everything you do downstream is wrong and then you present the wrong findings and then people make the wrong decisions so yes great data scientists are always very paranoid and putting sanity checks everywhere organizational skills yes

organization

this is by far my greatest weakness but what i've observed in great data scientists is that they are very organized they spend a lot of time thinking through their objectives or hypotheses and write down the steps that they will take to accomplish them and then actually follow through with it this is incredibly important because it's incredibly easy to go into analysis paralysis where if you're like me get distracted by all the cool things you can do that you lose track of the most important questions and objectives that initiated the project in the first place a great data scientist also organizes their code so this does not happen yeah what does um a stand for again storytelling you spend a lot of time on

storytelling

a project and you're really into it you find some really cool results or build an awesome model that solves like half the problems that people are always complaining about so you know you go and document everything and you show it to people super excited but guess what nobody cares everybody is too busy working on their own things and making sure that their own deadlines are met this is why storytelling is so important a great data scientist knows that the work is not done when they're done with the technical stuff they need to convey the awesomeness of their findings or what they made to non-technical people this is usually business people or product managers and you need to do it in a way that they can understand because you have to convince these people to actually use your work sounds crazy right but this is what i learned very quickly when i did my first analysis and i was super proud of it because i totally gave the team direction on what to do for the next project and nobody cared and it's totally my fault because it's not their job to decipher some obscure looking analysis i just plopped into their inbox so yes storytelling is very important both verbal and through writing and visuals bonus

BONUS: business domain

time great data scientists care about the business issues and their domain this one didn't make it onto the core list because i believe this is something that is more about finding the right domain for you to do data science as opposed to a skill in itself this is extremely important though because the data science role works very closely with business functions and leadership and those in are primarily concerned with wealthy business so you as a data scientist if you don't care about tech you would hate being a data scientist because who cares about new gadgets or services the team is building and you know if you don't like medicine why would you even bother to think about how to improve it you get me just to find the right business domain for you i think you really gotta put yourself out there i don't mean that you have to go get an internship or a job in like different domains which requires a lot of work and a lot of commitment instead you can totally do projects about domains that you think you might be interested in resources like kaggle have so many different data sets and the time commitment can be as long or as short as you like well i hope you enjoyed this video and it's helped you move forward in deciding data science right for you or what you can do to level up as a data scientist and finally don't forget to always minimize effort and maximize outcome i'll see you guys in the next video

Ещё от Tina Huang

How I would learn to code (if I could start over)

Tina Huang | 25.04.2021 | 13 сегм. | 3 449 134

How I consistently study with a full time job (9 months in and still going strong)

Tina Huang | 09.10.2021 | 5 сегм. | 2 888 742

How to learn to code FAST using ChatGPT (it's a game changer seriously)

Tina Huang | 02.03.2023 | 8 сегм. | 2 387 940

How to self study technical things

Tina Huang | 04.04.2021 | 10 сегм. | 2 322 419

Financial Literacy In 63 Minutes

Tina Huang | 01.01.2025 | 15 сегм. | 2 083 145

Google's 9 Hour AI Prompt Engineering Course In 20 Minutes

Tina Huang | 09.12.2024 | 10 сегм. | 1 552 364

Ctrl+V

Экстракт Знаний в Telegram

Транскрипты, идеи, методички — всё самое полезное из лучших YouTube-каналов.

Подписаться