Data Science for Real

(4-week online data science course)

In this course, you'll get the real raw data from a real online business. And you'll work on exactly the same data science tasks that you'd have to work on in real life.

Finally you'll:

  • get hands-on experience with data cleaning (and see how challenging it really is)
  • have to think about how your data science tasks can create actual value for a business (and go further than just writing the code they ask you to write)
  • pick and define the useful metrics by yourself (and learn how to ignore vanity numbers)
  • build statistical models on 100,000-line datasets (finally, you'll have enough data)
  • build statistical models on unprepared datasets (and see how unpredictable and uncertain real data can be)
  • experience all the ups and downs of a data science project (it's better to learn about these right now than on your first job)

"video placeholder"

The promise is simple:


Data science is hard.

This course will show you why.

And it will also show you how to overcome this.

This is more real than any other data science course out there.

other data science courses python machine learning

other data science courses...

data36 data science course data cleaning thinking automations

...vs. this course

How this course works.

It'll be pretty similar to the 6-week JDS course that you've taken already.

  • 1
    You'll get the raw data.
  • 2
    You'll get tasks that I'd give you if you were a junior data scientist and I were your manager.
  • 3
    You'll try to figure out the solutions by yourself (business decision, writing the code, choosing the right statistical methods, etc.)
  • 4
    Then I send you my solution -- video + code base -- and you can compare it to yours. (It’ll be just like you were sitting next to me while I’m doing a data science project for a real online business. I'll explain everything with full transparency: business decisions, statistical thinking and the code.)

But there are a few key differences compared to the 6-week JDS course, too.

  • 1
    This is not a simulation like Send-A-Tree was! You'll work with real data. It is not prepared -- so it'll be pretty messy. You'll have to clean it, figure out how you sort it into a useful data infrastructure and automate the whole process.
  • 2
    The business situation will be much more complex than it was at Send-A-Tree. Thus the problems and solutions will be less directed, too. The business thinking part will be more challenging. Choosing the right analytics methods will be less self-evident.
  • 3
    It's likely that there will be quite a few alternative solutions for the given tasks. And some of them will be worth further discussions and brainstorming sessions. (Hey, it can surely happen that you'll come up with a better solution than I do!) So I expect some interesting Slack group discussions. And if so, we can even set up a few Zoom calls to discuss the best alternatives.

All in all:


JDS and Send-A-Tree was like a real life data science project.


This course will be a real life data science project.

Will you be able to turn this...

raw data

...into this...

cleaned data in a dataframe

...then into this?

analysis on line chart

Or this?

clustering algorithm

Or into something even more exciting?

Curriculum - The agenda of the 4 weeks

If you enroll, you will receive a data science task every Monday, and Thursday for 4 weeks.


Two days after receiving each task you will get the solution from me: the code base sent in text format and the detailed explanation sent in video format.


The current plan for the curriculum (might change a bit):

WEEK #1

Monday

intro,

data cleaning

Thursday

preparing the data,

sorting it into tables (users, logs),

automations

WEEK #2

Monday

main business metrics

(funnels, KPIs, main dashboard),

what to focus on in later projects

Thurday

data discovery,

"best practice" analyses:

segmentations, best source report,

best performing articles report

WEEK #3

Monday

ad-hoc analysis #1

(conversion rate optimization:

what makes a good landing page)

Thursday

ad-hoc analysis #2

(retention analysis:

what makes users stick)

WEEK #4

Monday

Machine Learning.

Who’ll buy.

(Attribution modelling for source and articles.)

Thursday

Discover a new library!

(networkx)

WEEK #5

Monday

Closing, zoom call, discussions

Disclaimer: this is the beta version of the course

Let me be totally transparent about it: this is the first edition, so the beta version of the course. I know "beta" sounds lame - but in fact, for you it's going to be pretty cool. Let me explain why.


I’ve put a lot of energy and work (literally years) into creating the main concept, the curriculum and preparing the project… and during the course, I’ll also work hard to make everything just as high quality as possible.


“Beta,” in this case, really means that I’ll have much more personal involvement in this very first version of the course than I’ll have in the next iterations of it. For instance, I'll be pretty active on Slack and I will organize Zoom meetings when needed. Most of the videos will be also recorded throughout the course... Which on one hand means that they might not be "studio quality" (yet). But it also means that I'll be more flexible, I can react to your feedback and tweak the course into the direction that best serves the participants.


Beta means lower prices, too. I plan to price the final version of this course somewhere between $697 and $997. But this first launch will be available for a discounted $297 price.


You'll have lifetime access to the course and all future updates. Meaning: if there'll be a higher quality re-recorded version, you'll have access to that, too.


Also, there will be a Slack forum and priority email support for all participants!


And one last but important thing:

The course will be available only to former JDS students and only the first 10 registrants.

FAQs (well, probably you'd ask these...)

"What's the name of the business that we will work with?"

"Will you sell my solutions (code, idea, etc.) to a client?"

Presenter

Tomi Mester is a practicing data analyst and researcher for 7+ years.


He has worked for Prezi, iZettle and several smaller companies as an analyst/consultant.


He’s the author of the Data36 blog where he writes posts and tutorials on a weekly basis about data science, AB-testing, online research and data coding.


He's an O'Reilly author and presenter at TEDxYouth, Barcelona E-commerce Summit and Stockholm Analytics Day. More info about Tomi: check out the intro video. >>

Details

Start date: May, 2021


Price: $297 (plus your country’s VAT if you live in the EU)


Only for previous JDS students -- the max number of attendees is 10!


ps. If you are from Hungary, please email me before you register: tomimester@data36.com.

Guarantee

I worked (and will keep working) really hard to make this course the best available, and I stand behind it 100%.


I understand that enrolling in an online course is not always an easy decision, so I made this decision totally risk-free for you: if you request one, I’ll give you a full refund within the first 30 days.

Registration

Disclaimer!

After collecting feedback and data about this course, I realized that there's no significant demand for it in this format. At least, not for my audience. Hence, I decided to channel all the materials of this course into another one -- but that one is only available for the students who finished my flagship course already: The Junior Data Scientist's First Month. If you did finish that course and curious about this other one, please email me and I'll tell more. If you didn't finish JDS yet, go check that out first. Here>>