ZEMANTA
DATA SCIENCE
SUMMER SCHOOL

4TH EDITION | 11. – 15. JULY 2022

Zemanta Data Science
Summer School

Are you a student or a young professional, eager to learn more about data science and machine learning in the industry?

Do you code and enjoy analysing data?
Want to improve your practical machine learning skills?
Have you done some machine learning before and would like to know how we do in in the industry on a big scale?

Join our data science summer school! In one week we will go through all stages of data science projects – from defining the problem, gathering and analysing data, to building models, experimenting and analysing results.

We will work with gigabytes of realistic advertising data and solve real problems. A great opportunity for you to jump into data science head-first, gain some practical knowledge and experience, and see how we do it in the industry!

For a week you will be a guest in our offices, where our data scientists and engineers will help you use data science in practice. For this to work, bring solid coding skills and at least basic data science and machine learning knowledge.

While Zemanta Data Science Summer School is completely FREE of charge (including lunch, coffee, snacks and all activities), we do expect your full commitment during the one week of the summer school – remember: if you apply, get accepted and don’t show up, you are taking away a seat from someone else who really wanted to attend!

Who will benefit most from attending the Data Science Summer School?

Eager individuals who know how to write code (most of the summer school is very hands-on), have some machine learning and data science knowledge (eg. participated in a Kaggle.com challenge, experimented with some ML algorithms, took a ML course at university or online, etc), for example:

Final year undergraduate; masters or PhD students at Computer Science or related university programmes
Recent graduates from Computer Science or related university courses
Professionals in Computer Science or Data Analytics jobs, wanting to learn more about ML and DataSci in the industry

HOW MUCH DATA SCIENCE / ML EXPERIENCE DO YOU NEED?

You will learn a lot about ML and DataSci, especially in the context of how we do them in the industry. Having some experience and knowledge (eg. finishing some courses at university or online, or participated on a kaggle.com challenge) is preferable, you need to know at least the basics. Good rule of thumb is that if you ever fitted some models and made some predictions with them, then you probably know enough – summer school is set up in a way that DataSci/ML novices will learn the basics and do some hands-on experimentation with guidance, while experts will be able to really utilize their knowledge and do some additional magic on really cool data!

HOW MUCH PROGRAMMING KNOWLEDGE / EXPERIENCE DO YOU NEED?

Summer school is very hands-on, so knowing how to write code is crucial. We will teach you some TensorFlow with Python, but other than that, choice of favourite language is up to you! Good rule of thumb is if you did any studies at a computer science university programme or work as a software engineer, than you’re good!

Previous Summer Schools

1st Edition

2nd Edition

3rd Edition

What is the plan?

A week of incredibly interesting data science talks and workshops with a lot of hands-on experimentation – all guided by our expert team of data scientists and engineers! There will also be external invited lectures and a lot of interesting social activities with your fellow data scientists!

Programme 2022:

MONDAY

09:30 – 10:15 COFFEE & GETTING TO KNOW EACH OTHER
10:15 – 10:45 INTRO TO SUMMER SCHOOL
10:45 – 12:00 TALK: ONLINE ADVERTISING TECHNOLOGY – WHAT HAPPENS BEHIND THE SCENES
12:00 – 13:00 LUNCH
13:00 – 14:00 TALK: MACHINE LEARNING IN PRODUCTION AT SCALE
14:00 – 17:00 TALK & WORKSHOP: DATA SCIENCE WORKFLOW & TOOLS + DEV ENVIRONMENT SETUP
17:00 → PIZZA, BEER & GAMES

TUESDAY

09:30 – 09:45 COFFEE & CHAT
09:45 – 12:00 TALK & WORKSHOP: MACHINE LEARNING WITH TENSORFLOW
12:00 – 13:00 LUNCH
13:00 – 13:45 TALK & WORKSHOP: MACHINE LEARNING WITH TENSORFLOW – CONTINUED
13:45 – 16:45 WORKSHOP: FIRST HANDS-ON DATA SCIENCE CHALLENGE
16:45 – 18:00 SECRET GUEST TALK
18:00 → MINI GOLF IN TIVOLI

WEDNESDAY

09:30 – 09:45 COFFEE & CHAT
09:45 – 10:15 TALK: AUTOML
10:15 – 11:00 TALK: MULTI-TASK PREDICTION ON REALISTIC ADVERTISING DATA
11:00 – 12:00 DATA SCIENCE GUILD LEARNING FORUM EXTERNAL TALK
12:00 – 13:00 LUNCH
13:00 – 15:00 GRAND CHALLENGE INTRO & BRAINSTORMING
15:00 → GRAND CHALLENGE HANDS-ON
17:30 SNACKS & REFRESHMENTS

THURSDAY

09:30 – 09:45 COFFEE & CHAT
09:45 – 13:00 GRAND CHALLENGE HANDS-ON CONTINUED
13:00 – 14:00 LUNCH
14:00 – 15:00 ZEMANTA INVITED GUEST TECH TALK
15:00 – 17:30 GRAND CHALLENGE HANDS-ON CONTINUED
17:30 → ESCAPE ROOM

FRIDAY

09:30 – 09:45 COFFEE & CHAT
09:45 – 11:00 GRAND CHALLENGE HANDS-ON & PRESENTATION PREP
11:00 – 12:00 TALK: MULTI-ARM BANDITS & A/B TESTING
12:00 – 13:00 LUNCH
13:00 – 14:00 LIGHTNING TALKS
14:00 – 15:30 GRAND CHALLENGE SHOWCASE & KNOWLEDGE SHARING
15:30 – 16:00 GRADUATION CEREMONY

Why learn from us?

At Zemanta, we’re building the most advanced native advertising platform in the world.

HERE’S WHAT OUR WORLD-CLASS DATA SCIENCE TEAM IS WORKING WITH

over 500 million machine learning predictions calculated – every second
ML models we ship to production have an immediate effect on the business
we don’t keep cutting-edge ML and deep learning algorithms in the drawer – we deploy them to production
terabytes of data generated daily, stored on S3 and BigQuery
multi-cloud environment – AWS, GCS, private cloud
Go-lang is our main production language, python for offline research
production ML algos developed in-house in either Go-lang or TensorFlow
in-house distributed AutoML framework for effective offline research
powerful research server cluster
full ownership of the ML-powered features we develop – from idea to final product
a lot of learning – both internally (knowledge sharing initiatives) and externally (going to the best conferences, reading papers, watching talks)
sharing knowledge externally (talk at conferences and meet-ups, mentor students at DataScience@UL-FRI, Zemanta Data Science Summer School, …)
a part of a world-class data science and software engineering organization distributed across Zemanta and Outbrain

Zemanta joined Outbrain, the largest native advertising company in the world, and together we are on a mission of enabling everyone to discover great content.