Data Science Economy: Dawn of AI
2+ keynotes / 30+ talks
delivered from experts & practitioners

16th & 17th of May 2019, @KRAŠ Auditorium, Zagreb, Croatia
Join the


Workshop and hands-on presentation on the second day of the conference are a great format for participants to dive deeper into specific tools and products used within the Data Science field.
Working in groups and guided by the expert in specific filed of Data Science they will be introduced with basic principles, concepts, and techniques used for big data and business analytics. Participants will get a good picture of all these concepts and how they all are interconnected to each other in organizational context.







Marko Štajcer, Director of Innovation & Development, Poslovna Inteligencij

Machine Learning for the Elastic stack

Complex, fast-moving datasets make it nearly impossible to spot infrastructure problems, intruders, or business issues as they happen using rules or humans looking at dashboards. Elastic machine learning features automatically model the behavior of your Elasticsearch data — trends, periodicity, and more — in real time to identify issues faster, streamline root cause analysis, and reduce false positives.

With machine learning, users can go deeper and ask questions like “Have any of my services changed behaviour?” or “Are there any unusual processes running on my hosts?” These questions require behavioural models of hosts or services that can be automatically built from data using machine learning techniques. Elastic machine learning features make all that possible and apply to a broad range of use cases and datasets, allowing users to get creative with where and how you use them.

Marko Dobrinić, Data Scientist, Megatrend, Croatia

Developing real world applications with computer vision

IBM PowerAI Vision is a tool that enables quick labeling of data, as well as development and deployment of deep learning models. It includes the most popular deep learning frameworks and their dependencies, and it is built for easy and rapid deployment and increased team productivity. By combining PowerAI Vision software with accelerated IBM® Power Systems™, enterprises can rapidly deploy a fully optimized and supported platform with blazing performance.
In this workshop we will be covering some basic concepts of Deep Learning and diving into real world application with object detection and image classification use cases with IBM’s Power AI Vision.

Vladimir Marković, IT Auditor, Banca Intesa Beograd, Serbia

How unsupervised machine learning help in fraud detection?

How unsupervised machine learning help in fraud detection?
Introduction to internal fraud – some cases
SAS Visual Investigator demo
How to use advanced analytics in reducing sample of suspicious
– About anomalies
– Assumptions – choice of the method
– Some techniques in detecting anomalies
Steps in detecting collective anomalies using application logs in practice
– Data preparation
– Find behavioral patterns
– Make „fingerprint“ for each user
– Analyze the fingerprints and detect collective anomalies
Visualization of results
Some tips
Using results in SAS Fraud Solution

Neven Trgovac, Data Scientist, KOIOS

Scraping and analysis of the EU Parliament open data

In this workshop we will show the Python packages for data scraping, analysis and visualization we used in building our EU Parliament activity research. We will go through the most interesting and challenging tasks we faced, as well as some of the mistakes we made during analysis and implementation, by displaying the code in a Jupyter Notebook environment. We will also refer to the projects and materials we used and conclude by inviting further discussion and collaboration on our publicly available project.

Dario Radečić, Student Data Science, Algebra/Data Analytics, NEOS

Real estate market price prediction tool

Fast recovery of real estate sector in Croatia in last months is powered by tourism and expected Croatian EU presidency. No matter of you are looking to buy or rent, to live in or to invest, real estate is one of the preferred investor strategies. With upcoming law related to property tax, our market can expect changes both in way how price will be calculated but also in way how different parts of Country will be further developed. With recent Airbnb ban from Paris/France aiming against real estate pricing trends, we can expect more and more policies to enter the market.
If you wonder why data scientists are not covering this topic – wonder no more! As part of Data Science master program on Algebra University College in co-operation with data science team from Algebra Lab, session will introduce data science based smart real estate market prediction tool. Let’s make future together – light your spark

Badr Ouali, Data Scientist, Vertica

Workshop on 3 use-cases showing the power of Vertica advanced analytics

• Introduction to Vertica-ML
• Geospatial + ML to predict golf players location and apply a marketing policy in their location (I explain what is possible with the geospatial library first and then I go to the use-case) – 30mns
• ML to catch fraudulent machines which are clicking a lot in a PPC system – 20mns
• Guess: Predicting clothes generation in order to segment customers – 20mns

Arian Stipić & Tomislav Bronzin, CITUS

I see you, I know who you are, what you want, and you cannot fool me!

There are many ideas about incorporating augmented reality in everyday applicative solutions, but very few practical examples currently exist. Most of them are primarily related to gaming industry and this workshop will show you how to implement real-life solutions based on this very advanced technology.

The solutions are built on Microsoft Kinect device support for stereoscopic real-time image recognition and analysis. These Kinect capabilities are then combined and integrated with other components in order to create lightning fast analysis and augmented reality experience for the end user.
Attendees will get the insight into the whole development process, see the practical demos and have the opportunity to discuss all aspects of presented solutions.