agenda
14, 15, 16 November 2019
9:00 - 10:00
Registration / Networking
10:00 - 10:20
Main Stage
Opening
10:20 - 10:40
Main Stage
Debunking preconceptions about data collection: it's hard, it's slow, it's costly
Laurent Nguyen
Data Scientist, Eyeware
Data quality is essential to building high quality AI pipelines. Often, the processes of
collecting, annotating, and cleaning data is often regarded as the poor relative of data science despite
their importance and the cost in time and money they require. In the many cases for which data is not
available, implementing methodologies to efficiently collect and prepare data can strongly speed up data
science projects. In this talk, I will provide some insights about my experience in performing these
tasks.
10:40 - 11:00
Main Stage
Towards Autonomous Automotive Engineering - How AI changes mechanical engineering and manufacturing
Christian Drescher
Member of Artificial Intelligence Research group, Mercedes-Benz
Automotive development is an expensive and time-consuming task. Artificial Intelligence
can help to speed it up, e.g., through powerful machine learning models that approximate the processes
involved in mechanical engineering and manufacturing. This talk will showcase potential future
applications of AI in the automotive domain.
10:00 - 11:00
The EU for Civil Society Activists Room
Introduction to Open Source (Part I)
Lukas Andriukaitis
Associate Director, Digital Forensic Research Lab
Today, people are more interconnected than at any other point in history. The Atlantic
Council’s Digital Forensic Research Lab (DFRLab) is building a leading hub of digital forensic analysts
tracking events in governance, technology, security, and where each intersect as they occur. This
presentation will demonstrate the power of open source analysis. DFRLab will demonstrate how they are
using open source intelligence to help identify, expose and explain disinformation and malicious
narratives online. In this presentation you will learn about the 4 D’s of disinformation, the method of
geolocation, online source verification and will have an opportunity to try some of the methods yourself.
10:00 - 10:20
ODIHR Human Rights Room
Use the Voronoi
Harry Stevens
Visual Journalist, Washington Post
In addition to being beautiful, Voronoi diagrams can be extremely useful when designing data
visualizations. This workshop will show how to create and use Voronoi diagrams using D3.js.
You will need a laptop as well as basic knowledge of JavaScript and D3.js.
10:00 - 11:00
ForSet DataViz Room
dActivism (data+activism workshop) (Part I)
Market Cafe Magazine
Co-founders, Market Cafe Magazine
During the workshop Market Cafe Magazine founders Tiziana and Piero will walk you through
the process to create a thought-provoking data visualisation protest sign following their signature 4D
creative process: Discover, Define and Design with Data. The workshop will be a balanced combination of
inspirational talks and fast-paced interactive exercises, to help you understand the key phases of
designing with data, from the definition of goals, to the production of the final outcome.
10:00 - 11:00
BoG Fintech Room
Chatbots that knows if you are angry
Cheuk Ting Ho
Data scientist, Santander
While chatbots are a blooming thing, we always want out bots to be smarter. Imagine a
chatbot that knows how angry your customer is and handle the complain more seriously. With the development
of NLP, seems we can get more out of the text, like extracting the tone and the emotion form the text.
For this workshop yu will need: a laptop, basic Python coding experience. Python >= 3.5, an IDE of your choice (details can refer to setup section in the workshop: https://github.com/Cheukting/rasa_workshop#install-rasa-and-set-up-the-environment)
10:00 - 11:00
Maxin AI Room
Machine Learning in R (Part I)
Chris Kennedy & Nutsa Abazadze
Data Scientist, Berkeley Institute for Data Science
This workshop will introduce the fundamentals of applied machine learning using R. Using
interactive R Markdown notebooks, you will explore key procedures like data preparation and
cross-validation, along with intuition and sample code for the most popular machine learning algorithms:
lasso, decision trees, random forest, xgboost, and SuperLearner ensembles. You will be empowered to apply
machine learning to your own data and will have resources for self-study.
You will need a laptop with R 3.5+, RStudio. You ashould also have an understanding of basic R, including installing packages and manipulating data frames.
11:00 - 11:30
Main Stage
Improved Sentinel-2 image registration or is something wrong with EU Copernicus data?
Nikita Zhuravlev
The Copernicus Sentinel-2 mission comprises two satellites placed in the same
sun-synchronous orbit that monitor variability in land surface conditions. The satellites provide a lot of
data, which helps us to detect field boundaries, present up-to-date information on field conditions, as
well as predict crops and sowing dates.
The challenge lies in pre-processing this data. This is because the imagery of any given region
experiences spatial shifting due to a host of various factors. Shifts can reach up to 100 meters, which
makes it hard to determine field boundaries with precision.
I'll talk about how we solved this problem by coregistering imagery. I'll also cover how we overcame the
problem of cloud cover, snow cover, and the images' enormous size, which presented us with an obstacle in
terms of processing massive amounts of data received from Earth.
11:00 - 11:30
Break
11:30 - 12:00
Main Stage
Visualizing mobility
Ilya Boyandin
Data Visualization Engineer, Teralytics
Understanding human mobility is important for fields like urban and transportation
planning, migration studies, epidemiology, disaster response. Data on human mobility can be complex and
difficult to interpret in tabular form, therefore visualization plays an important role in their analysis.
In this talk we will discuss the challenges of the analysis of mobility data, the modern visualization
approaches and the tools available today.
12:00 - 12:30
Main Stage
Deep learning for sensor-based activity recognition
Andjela Todorovic
Software Engineer @ Cubic Corporation
Human activity detection has seen immense growth in the last decade playing an important
role in the field of pervasive computing. This popularity can be credited to its real-life applications
primarily dealing with human-centric problems in general medicine. Many research attempts with data mining
and machine learning techniques have been undergoing to accurately detect human activities for healthcare
systems. I would review some of the predictive machine learning algorithms and compare the accuracy and
performances of these models.
11:30 - 12:30
The EU for Civil Society Activists Room
Introduction to Open Source (Part II)
Lukas Andriukaitis
Associate Director, Digital Forensic Research Lab
Today, people are more interconnected than at any other point in history. The Atlantic
Council’s Digital Forensic Research Lab (DFRLab) is building a leading hub of digital forensic analysts
tracking events in governance, technology, security, and where each intersect as they occur. This
presentation will demonstrate the power of open source analysis. DFRLab will demonstrate how they are
using open source intelligence to help identify, expose and explain disinformation and malicious
narratives online. In this presentation you will learn about the 4 D’s of disinformation, the method of
geolocation, online source verification and will have an opportunity to try some of the methods yourself.
11:30 - 12:30
ODIHR Human Rights Room
Data clinic: how to use data and new technologies for your human rights investigations
Milena Marin
Senior Advisor for Tactical Research, Amnesty International
Bring your own data – in this workshop participants bring troublesome data and human rights investigations
and the facilitator, together with other participants troubleshoot and advise how to work with it.
This is a private session not open to general public
11:30 - 12:30
ForSet DataViz Room
dActivism (data+activism workshop) (Part II)
Market Cafe Magazine
Co-founders, Market Cafe Magazine
During the workshop Market Cafe Magazine founders Tiziana and Piero will walk you through
the process to create a thought-provoking data visualisation protest sign following their signature 4D
creative process: Discover, Define and Design with Data. The workshop will be a balanced combination of
inspirational talks and fast-paced interactive exercises, to help you understand the key phases of
designing with data, from the definition of goals, to the production of the final outcome.
11:30 - 12:30
BoG Fintech Room
When big data goes fast - real-time data stream processing
George Mamaladze
Senior Software Architect, Siemens Corporate Technologies
After you successfully mastered batch processing your data, get ready for the data stream
processing! This is when instead of capturing data and then running batch jobs on it, processing is done
as the data arrives to extract valuable information more quickly. In this workshop, George will share his
experience with large scale data stream processing architectures, very specific architectural patterns and
techniques that have emerged for these systems in the industry. He will also touch very unique challenges
of building such systems and try to understand the core differences between traditional and real-time data
stream processing applications.
This workshop requires expertise in IT and Software Development is preferable
11:30 - 12:30
Maxin AI Room
Machine Learning in R (Part II)
Chris Kennedy & Nutsa Abazadze
Data Scientist, Berkeley Institute for Data Science
This workshop will introduce the fundamentals of applied machine learning using R. Using
interactive R Markdown notebooks, you will explore key procedures like data preparation and
cross-validation, along with intuition and sample code for the most popular machine learning algorithms:
lasso, decision trees, random forest, xgboost, and SuperLearner ensembles. You will be empowered to apply
machine learning to your own data and will have resources for self-study.
You will need a laptop with R 3.5+, RStudio. You ashould also have an understanding of basic R, including installing packages and manipulating data frames.
12:30 - 14:00
Networking Break
14:00 - 14:20
Main Stage
An intorduction to AI
Tom Hunter Smith
Senior Data Scientist, LoopMe
A brief introduction into Artificial Intelligence, its history, the common misconceptions
surrounding it and in particular how it can be used effectively to optimise the delivery of advertising
campaigns in the digital ecosystem.
14:20 - 14:40
Main Stage
Dark Archives for Machine Learning
Liviu Pop
Researcher, Romanian Academy
Dark archives usually refers to archives that are inaccessible, that are hidden from view.
The second half of the XXth century saw a number of political systems that left behind all kind of traces.
If we were to use the archives of that period as a source of data for machine learning algorithms, which
caveats should be taken into consideration? Not only those archives are hidden, but some of them are
hiding dark stories that deserve to be retold carefully. This talk raises some questions and offers a few
answers on how to approach those archives.
14:40 - 15:00
Main Stage
Change the game - Create a new world with a simple data
Salome Urotadze
AI Developer, SpaceX, Neuralink and Microsoft
It’s already known that we can change the world with the latest technologies, but what if we create a new
one to help the whole universe in evolution? Every little detail matters, let’s find out how can we use
the knowledge which is around us.
15:00 - 15:20
Main Stage
Why it’s not enough to be a good programmer
George Mamaladze
Senior Software Architect, Siemens Corporate Technologies
In the modern world of emerging technologies, we as humans need to be ready for a race
against machines by constantly extending our skillset. In this George will share his thoughts and
experience about why it’s important to acquire cross-domain expertise for every programmer to cope with
upcoming challenges.
14:00 - 15:20
The EU for Civil Society Activists Room
Open Data to Achieve the Sustainable Development Goals
Lorena Rivero del Paso
Manager for Technical Collaboration and Cooperation, Global Initiative for Fiscal Transparency
Lorena will reflect on the role of data for decision-making regarding the Sustainable
Development Goals. What type of data we need, how and possible uses of data science to improve policy
decision making.
14:00 - 15:20
ODIHR Human Rights Room
Let's make a map!
Roman Sverdan
Data Visualization Designer, Engineer | DW Dataship Fellow
Interactive maps are a great tool for advanced data visualizations but too complicated to
create. Do you think that way? Take your laptop, join the workshop and you will never think the same.
The requirements for this workshop are: a laptop, basic understanding of HTML and CSS
14:00 - 15:20
ForSet DataViz Room
Intro into machine learning (Part I)
Zdenek Hynek
Developer, Data4Change
Want to teach a computer what a fake Instagram profile looks like? Or how to tell if
humans are angry? Using Data Science Studio software, you’ll be able to try out some basic machine
learning methods. Please follow the instructions at https://introducing.ai/instructions/ to setup the
software required for the workshop
You'll need to have a laptop for this workshop. Also, you'll make your life easier if you install Data Science Studio (available at https://www.dataiku.com) beforehand.
14:00 - 15:20
BoG Fintech Room
Tableau in Action: Become a data champion at your organisation
Beso Elbakidze
Visual Analytics, Deloitte
This hands-on workshop will teach you how to analyse and visualise data with Tableau, and
by the end of the workshop, you'll be able to turn data into insights and continue your data analytics
journey on your own at your organisation
For this workshop you will needa laptop and Tableau Desktop.
14:00 - 15:20
Maxin AI Room
Machine Learning in R (Part III)
Chris Kennedy & Nutsa Abazadze
Data Scientist, Berkeley Institute for Data Science
This workshop will introduce the fundamentals of applied machine learning using R. Using
interactive R Markdown notebooks, you will explore key procedures like data preparation and
cross-validation, along with intuition and sample code for the most popular machine learning algorithms:
lasso, decision trees, random forest, xgboost, and SuperLearner ensembles. You will be empowered to apply
machine learning to your own data and will have resources for self-study.
You will need a laptop with R 3.5+, RStudio. You ashould also have an understanding of basic R, including installing packages and manipulating data frames.
15:20 - 16:00
Break
16:00 - 16:20
Main Stage
Electing Women in 2020: EMILY’s List Data and Technology
Mike Sager
Chief Technology Officer, EMILY’s List
After the 2016 election, over 40,000 American women told EMILY’s List they were interested
in potentially running for office. EMILY’s List Chief Technology Officer Mike Sager is part of the team
that worked to build the infrastructure to handle this influx while also overhauling the data and
technology program at the established organization. Mike will walk through how EMILY’s List handles data,
and provides tools and support to the Democratic women who are running for office, or considering a run.
16:20 - 16:40
Main Stage
Open Data: Research vs Privacy
Elena Poughia
Managing Director, Dataconomy Media & Data Natives
In recent years more and more governments and enterprises have been opening up data they
aggregate. According to the open data index published by the Open Knowledge Foundation, Taiwan is leading
the list followed by the UK and US. Germany's open data rank is currently only 24. As we are entering the
data-driven future, organizations and governments realize that opening up data is less of a choice and
more of a necessity. While opening up data can significantly enhance efficiency, raise trust and add
social value, we are facing a set of new challenges and risks than ever before - starting from the ethical
issues in AI and ending with the absence of data-driven education. This talk will focus on the current
state of open data, touch upon the importance and challenges of data ethics and offer entry points for
institutions for a critical assessment of their data policy as well as guidelines on how to open up data
consciously and consistently.
16:40 - 17:00
Main Stage
Can you hear your data talking for you?
Paul Imre
Founder, IMRE
Paul will let you explore how your data will talk for you in the near future. He will
start on a journey that will take you from inbound marketing to conversational AI and then finally to the
concept of marketing to machines. Are you ready to be out of the loop?
16:00 - 17:00
The EU for Civil Society Activists Room
Visualising topics emerging from academic publications
Antonio Campello
Data scientist, Wellcome Trust & Datakind
The ability to identify emerging topics from the academic literature is crucial in order
for funders to ensure appropriate coverage and appropriate focus. At the same time, researchers often want
to know how their contributions fit into the broader research landscape, and policy makers want to search
for relevant literature to substantiate their claims. In this workshop, Antonio will discuss how to assist
with some of these questions through open data and machine learning techniques. In particular, he will
present various data-driven methods to visualise research topics emerging from academic publications. He
will explain new data dimensionality reduction techniques (such as parallel t-SNE and uMAP), as well as
topic similarity concepts, and how to incorporate domain knowledge to assist the process.
16:00 - 16:20
ODIHR Human Rights Room
How to communicate budget data
Tarick Gracida
Technology & Communications Coordinator, Global Initiative for Fiscal Transparency
Communicating budget data to citizens is one of today's biggest challenges for governments
and civil society organizations around the world, in societies less patient in facing a lot of information
in their daily lives. This workshop presents tips to generate a comprehensive communication strategy,
online + offline, beyond visualizations and "simplified" versions of the data.
16:00 - 17:00
ForSet DataViz Room
Intro into machine learning (Part II)
Zdenek Hynek
Developer, Data4Change
Want to teach a computer what a fake Instagram profile looks like? Or how to tell if
humans are angry? Using Data Science Studio software, you’ll be able to try out some basic machine
learning methods. Please follow the instructions at https://introducing.ai/instructions/ to setup the
software required for the workshop
You'll need to have a laptop for this workshop. Also, you'll make your life easier if you install Data Science Studio (available at https://www.dataiku.com) beforehand.
16:00 - 17:00
BoG Fintech Room
Agile Data Science - Challenges and Upsides
Irakli Gogatishvili
Head of Data Analytics & Product Owner, Bank of Georgia
Agile transformation of data science teams entails challenges setting them apart from
software development realm. In this workshop Irakli will discuss upsides of the process with
methodological and practical implications of successful implementation.
16:00 - 17:00
Maxin AI Room
Machine Learning in R (Part IV)
Chris Kennedy & Nutsa Abazadze
Data Scientist, Berkeley Institute for Data Science
This workshop will introduce the fundamentals of applied machine learning using R. Using interactive R
Markdown notebooks, you will explore key procedures like data preparation and cross-validation, along with
intuition and sample code for the most popular machine learning algorithms: lasso, decision trees, random
forest, xgboost, and SuperLearner ensembles. You will be empowered to apply machine learning to your own
data and will have resources for self-study.
You will need a laptop with R 3.5+, RStudio. You ashould also have an understanding of basic R, including installing packages and manipulating data frames.
17:00 - 17:40
Main Stage
Panel
17:40 - 18:00
Main Stage
Closing
18:00 - 20:00