Tetiana Ivanova: How to become a Data Scientist in 6 months | PyData London 2016
Вставка
- Опубліковано 23 лип 2024
- Tetiana Ivanova: How to become a Data Scientist in 6 Months, a Hacker's Approach to Career Planning
PyData London 2016
This talk outlines my journey from complete novice to machine learning practitioner. It started in November 2015 when I left my job as a project manager, and by April 2016 I was hired as a Data Scientist by a startup developing bleeding edge deep learning algorithms for medical imagery processing.
SHORT INTRO
Who I am, my background and a short summary of my story. Here I will list the steps I personally took to achieve the goal I had.
HOW DID I DO IT?
Why I chose a “hacky” way to enter this career path. The first mover advantage, why getting a degree doesn’t always improve your career prospects. Possibly a rant on the signaling function of formal education and how that is rarely aligned with a relevant practical skill set. Some stats to back it up (best career success predictors). Examples of hacking bureaucracies/social hierarchies from my experience and elsewhere.
List of things not to do and common cognitive pitfalls.
Networking for nerds - how to do it right.
Time management for chronic procrastinators - how to plan a self-guided project. Some notes on psychology of time discounting and need for external reinforcement, with autobiographical examples.
CONCLUSION
You don’t need a Ph.D. or even a master's to do machine learning. On taking calculated risks and especially calculated exits from one’s comfort zone. Some notes on soul searching and how to choose a career that is also a passion. Reading list.
Slides available here: www.slideshare.net/TetianaIva...
www.pydata.org
PyData is an educational program of NumFOCUS, a 501(c)3 non-profit organization in the United States. PyData provides a forum for the international community of users and developers of data analysis tools to share ideas and learn from each other. The global PyData network promotes discussion of best practices, new approaches, and emerging technologies for data management, processing, analytics, and visualization. PyData communities approach data science using many languages, including (but not limited to) Python, Julia, and R.
PyData conferences aim to be accessible and community-driven, with novice to advanced level presentations. PyData tutorials and talks bring attendees the latest project features along with cutting-edge use cases.
0:00 Introduction
4:56 Higher education
13:02 Things not to do
17:55 What did I do in the end?
21:51 Time discounting and willpower
26:27 Time management techniques
31:37 Networking
37:18 Resources for Data Science transition
44:34 Don’t get started
45:51 Q&A
S/o to github.com/anzelpwj for the video timestamps!
Want to help add timestamps to our UA-cam videos to help with discoverability? Find out more here: github.com/numfocus/UA-camVi... - Наука та технологія
WOW! One of the most EXCEPTIONALLY compelling videos on UA-cam!!!
Listen to her! These are absolutely essential skills for LEARNING HOW TO LEARN ... not just for Data Science ... but for ALL disciplines, for all examples of lifelong continual learning. Master these skills -- you will need to use them throughout the next six or seven decades of your career.
I watched this 5 years ago as a student. I’ve been a Data scientist/ML engineer for the last 3.5yrs. Time flies. I remember how i felt that it would be a dream to do this for a living. For everyone chasing a dream, don’t stop the hustling. It’s going to happen 🙂
Any aspiring data scientists or ML engineers. Feel free to ask me any questions in the comments. Happy to contribute back!!
What field you are doing in data science?
@@wongtungtung7456 telco industry mostly
Am lost, Please help! As I have math background I am trying to learn one of this [Data science, Deep learning]. but I don't know where to start! I already have learned the basics of programming. I would like to build something maybe with Google Colab Notebook!
@@IsaacAdam0 ok first follow the two links fully and then do link 3
1. ua-cam.com/video/ua-CiDNNj30/v-deo.html
2.ua-cam.com/video/LHBE6Q9XlzI/v-deo.html
@@IsaacAdam0 finally link 3 ua-cam.com/video/JwSS70SZdyM/v-deo.html
This talk is even more relevant 7 years after.
I will come back to this video after 6 months to report my progress upon adopting and applying some of the wise recommendation from this talk
Such practical look at human psychology.
I have degree in math, got it 2 years ago, the rest could be skipped.
Two observations. Not every pretty girl is on tiktok though they make more money, much more money, than a data scientist could ever imagine. Secondly Tetiana you should setup your own Data Science school.
Why do we do our degrees? Perhaps it's about getting a formalized education? I did not get a degree to have someone devalue it by calling it a Prada bag. If you did not want to get a degree fine, you don't have to attempt to devalue it for others. And it is most likely in some way due to your degree that you were able to self-teach or whatever you did to become a data scientist. I am also a mathematician who has transitioned to being a data scientist and I have taken some courses and my classmates really struggled with many mathematical concepts and thinking that were taught to me in my college education.
Warning: this has NOTHING to do with data science. This is just one person airing their grievances. Plus a bunch of business buzzwords, half-truths and bs (starring Malcolm Gladwell, of course). Avoid.
Yeah, PyData it's the equivalent of TEDx, some are good, some are a waste of bandwidth.
This is a very informative piece. I'm a college senior starting my 5th year as a Comp sci major. I only just recently decided I wanted to focus on machine learning yet felt overwhelmed because I had no idea how to get started. This woman right here.....Thank you!
Thank you - this is incredibly valuable for me as I start on my journey to being a Data Scientist.
Excellent presentation, You gotta love Tetiana, very well presented obvious conclusions of higher education. Thumbs UP!
Thanks Tetiana, this has been a very enlightening talk. We generally don't get a lot of discussion around how to manage your focus and timeline when you are working alone after dropping out. A very to the point presentation on how to kickstart your data science career through non-conventional route of self learning.
Noted every single point of this presentation. It was so simple to understand. Thank you very much Tanya Ivanova. Lucky that I got through your presentation.
One of the most encouraging video I ever watched...Really Awesome!!!
Coursera Learning How to Learn is the first step for those, who's gonna take the self-study-marathon!
Thx for your Story and Presentation =)
wonderful insights... hat's off... that was a really great talk. amazing lecture!!!
This was a really awesome presentation, thank you so much! Currently in the best job of my "career" (Project Management/Computer Engineering) and have no college degree to show for it.
Wow thank you for sharing, I think more people need this perspective!
Insightful presentation, and I liked the narrative describing the approach she took in the career transition
This is an amazing speech and helped me a lot ! Thanks for deliver this amazing speech.
Oh my god one of the best presentations about how to start!
You can say that again! Seems so realistic and practical.
Molodec Tania! Well done Tania :) I'm currently into machine learning & data science. I don't do it for career purposes but build the service using these techniques
This is so helpful, such a great talk! I'm going through the same situation you were. I want to update my knowledge in communication strategy and it has been so difficult to find a course/MA that covers all the areas I want to learn. Communication has changed soo much in the past years... So scoping my own area of study is a very good solution. Thank you so much for showing me the way to start it! Cheers from Brazil.
Excellent talk, Tetiana. Thanks for sharing your experience.
Really a great talk! Loved it and I agree with it 100%
This was amazing .learned alot. and now gotta get this DATA science Jazz
great talk! very inspiration and informative. well done and good luck
This speech is an inspiration to me. I graduated with a useless liberal arts BA and I work in customer service (luckily, I'm not an american so I'm not debt-ridden) and a couple of weeks ago I decided I would become a data scientist. After some reading I discovered there's so much prospect in this profession and so much info on the net that it's a wonder why anyone bothers going to college anymore. And seeing that there are so many people who are doing this by themselves is really encouraging.
Tremendously inspiring talk to me. (I have a PhD btw.)
Having a PhD and postdoc doesn't make it easier to find a industry job. If anything, it makes it harder!
In my opinion everyone accustomed to problem-solving is a "scientist".
? Can anyone provide another link for the presentation
link is not working.
www.slideshare.net/TetianaIvanova2/how-to-become-a-data-scientist-in-6-months Check here
Excellent talk (Inspiring).
Thanks Tanya :)
Really good presentation, so much value on it! Sadly in my contry education is just another business, so msc and phd are not an option for me.
Really great presentation, thanks!
Great Tips! Make sense and you are able to relate to as well!
What was the book by Jaynes?
How important is a math background? I am getting mixed information on this.
Great speech, thank you :)
Amazing! the guy seemed angry, hope he is ok, with a job, a happy with his knowledge
Please repeat the question ( ͡° ͜ʖ ͡°)
frozen mustang 🎹
ha ha, 37 people disliked it,so those are just 37 sample from the 824 population with only .044 probability so we are 99.99% confident that they do not belong to this population :-P
So insightful this. Deserves an audience well beyond Pythonistas.
The study quoted as a motivation for this was an observational one of multiple studies back to 1980. That's 36 years ago.
Association was "proven" through what appears to be a regression analysis of this data, which outside of coefficient tables, is not displayed in any other way.
It also can be inferred that there was severe multicollinearity present between variables since there is a correction weight for GMA. The speaker did not show it in her presentation but once weights are applied all the "low importance" one's become 0.5.
Lastly, the link is to a blog post on the paper!!!
Real paper is here : mavweb.mnsu.edu/howard/Schmidt%20and%20Hunter%201998%20Validity%20and%20Utility%20Psychological%20Bulletin.pdf
love this talk :)
math was not my fav subject in school or college, and i have no experience in coding or anything close. There are courses online for beginners that says coding is recommended. My questions is should i go this path? do i have a chance?
if you don't like math it may not be for you, but i would starting at codecademy.com and finish the python coding course. this gets your feet wet right away and python is widely used for data science. The hardest part of programming is the basics IMO. Then Udacity has a great data analyst nanodegree program. There are plenty of online resources as long as your interested and willing to struggle. This is how i started programming and currently what I am doing. Good luck!
Andrew Gill thank you!
The only required skill is a reservoir of curiosity. You will need some basic math, linear algebra and statistics which in my opinion are easier than calculus which is not required. Programming would be actually a fun part and there are tons of good material online (choose Python or R, other options are much harder). But remember curiosity and love of participating in data science projects are the only prerequisites indispensable for success.
many thanks for speech :)
Love it.
awesome presentation
10/10 for that video!
Great talk
Nice presentation.
Thinking of quitting my job to learn data science.
Me to. Want to collaborate?
Already quit my job.
What are you currently studying?
courses.edx.org/courses/BerkeleyX/CS188x_1/1T2013/info
You?
I'm following this frame work "github.com/datasciencemasters/go"
I've thought about doing a Udacity Nanodegree as well, but i'm not fully sure if I should make the investment or not. However, its really appealing that Google is involved. Do you have a github account?
Good lecture but be aware that it's not concretely on how to become a data scientist, it's how to become everything.
If you had money to support yourself for a year it should take into account looking for next job, so it was really not a full year as applying for a job is also time consuming.
good lecture, sadly I'm procrastinating right now
everyone face that dude, but dont lose the hope start new day with fresh things
she is really intelligent women
nice
Take a dance class!!! Never thought about humiliation vaccination! Brilliant
let's say ...6 years!
I can completely understand why people pursuing a Phd or have already gotten their Phd are getting defensive. It's at least 3 years of your life and none of you should be discredited in any way. Tetiana made sure to mention that this was the opinion of group sample=1, so please try to find it in yourself that there are some people who might not agree with some of the motivations for you taking a Phd. You Phd's are awesome and you are probably around a lot of people who support your endeavours, but I could imagine that you are not around a group of people who don't agree with your pursuit, this doesn't mean that a lot of people don't disagree so don't be surprised when you run in to people who don't, and please, please don't react like the lady did in the end, you are brighter than that (some of the brightest I might say!).
I like this video, but I feel like you should disclose the fact you were a math major in college. That's a huge advantage in trying to be a data scientist.
I think she did - At 1:08, she mentioned being a mathematician
Очень красивая и умная девочка . Будь по-прежнему мое сердце .
Also I'd say if you are just starting out your career direction go be an intern, anything, in a particular industry/area and see if you are really interested first. For instance if you want to be a lawyer you can become a paralegal first.
You want a CS degree and want to code?, start coding now.
Get a feel for industry, what the job(s) entails, then you'll know if you really want to pursue it and might see a good way to "hack" your way there faster..
I have found that when you are in line for the most senior role in a department the people interviewing you other than reading your resume about your accomplishments and checking out your credentials they have no way of determining your ability. If indeed you are at the level you are applying few if any in the company may be able to interview you and determine if you are bsing them. ok maybe not. What is true is at senior levels organizations are inundated with resumes the easiest way to knock X% of the resumes is through education credentials and certifications. sure maybe you miss out on some really good self made people but lets face it we are not a talking about Einstein material here. These are regular jobs. Actually, because data science is relatively new the ability to hack a career is possible since the demand probably outweighs the number of candidates with degrees so. she is in right career at the right time. try to hack into a CFO job or structural engineer and see how far that gets you without credentials.
Good Talk! But it did seem like she is overcompensating for not doing her PhD! You are doing good job .. Since you do not have a PhD, I don't think you'd know what doing PhD entails... Not all PhDs would provide visible contribution to the society but everything that led to the state of data science today was the endless efforts of many many PhD students and graduates! Your talk was good, you don't need to belittle PhD!
Not to be the negative here but a bsc in mathematics from UCL is strong enough to get you in...so please spare me the dropout style
many so-called data scientists can't find a job, let alone a lucrative one. This unique example is like dropping out of college to get rich just because Bill Gates did so .
Stop calling calling yourself 'Mathematician' .. you only have BSc degree. Second, you worked in sales, project - nothing to do with Data Science..The company that hired you seems like a startup, may be the Coursera helped. Calling onself 'Data Scientist' .. whilst calling others who've actually done PhD, published and developed those machine learning algorithms unworthy is = rude... Nice way to glamor by organizing meetups!
Whats your Kaggle ranking?
THANK YOU. Jesus, finally a voice of reason in this sea of sycophants.
Mind linking us to your :"How to become a Data Scientist " presentation ,please?
watch from 19:40
Everybody who don't have a PhD spend so much time talking it down. I smell envy.
Oh another fancy story how to become SCIENTISTS in 6 months) soooo keel) and than become Grand Master on topcoders in 3 weeks, just use time management, and believes in your self) positive thinking. Yes we can)
So funny how people love to believe in any easy way to succeed.
She is not saying or implying any of that she is just describing a path for being a data scientist without getting a degree
Tetiana, although I admire the way you present your matter I think that you should not use the term scientist but analyst. 6 months of self-learning surely doesn't make us (I do the same thing) scientists. And you can't just belittle a phd student because some of them, are not doing this for the right cause. 6 months or even 2 years dealing with a matter are not going to make someone scientist.
where do you draw the line between scientist and analyst?
is it experience? you make it sound so. just curious.
If you are a data scientist, what have you discovered ?
Example: Newton was a scientist, since he discovered the universal law of Gravity.
Lol.. True.
Ask her Kaggle ranking? and / or if she published any genuine work - in any area forget data science, even algebra...
I think a data scientist is a scientist not so much because of discovering new laws of nature but rather because of sticking to the scientific method of inquiry into things, data sets in this case.
Engineers also use scientific knowledge to build devices. They do not call themselves scientists though.
Your argument makes sense. But then it becomes just an argument about terminology, which does not seem that important.
On the other hand, producing knowledge from data sets can maybe later be generalized to some laws about data that we still do not know. As for engineers, they do not always use scientific knowledge and method, sometimes it is as much art and trial-and-error before they build some device or software product.
For a data scientist, it is important to not hold any bias, formulate hypotheses and then test them, rejecting if not enough evidence is found, which is very similar to the scientific method.
For example, statistics is a branch of science. It would be logical to conclude that a statistician is a scientist. But can we say that a statistician is not a scientist if he is working in applied statistics and not in "pure" research and doesn't produce any new knowledge? I think it's a difficult question to answer because it is highly subjective.
"Data Scientist" is a silly name. So it "Social Scientist". At least in the pure sense. They're not discovering universal truths about nature. And in a lot of cases, they're barely using the scientific method.
On the other hand, I'm using "data science" to analyze terabytes of atmospheric data to better understand ice crystal formation in clouds. And I'm not even "scientist" . . . I have a ph.d. in chemical engineering.
You are NOT a mathematician, please do not insult those who have rightly earned the privilege of calling themselves as such.
But she is, she has BSc in Mathematics from University College London. And she was humble throughout the speech, and she answered questions with scientific precision and intelligence that most of PhD's genuinely lack.
@@anismatar exactly. Not discuraging people and their studies but the approach.
Tetiana Ivanova, Trump won and i am a data scientist, you really shouldn't use Trump as an example, that is not funny
trump won!