Buy tickets

Sava Centar, Belgrade

20-24th November 2023

DSC SCHEDULE

Download Preliminary Schedule
DATA & AI TALKS
BREAKOUT SESSIONS
CO-LOCATED EVENTS

TECH-tutorials

Stream 1

Analyze your data in Python

Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
Tool Showcase
Beginner to Intermediate
Data Analytics is one of the most important skills, which is of key importance for the successful operation of any Data Science team. Especially if the data analysis is done in one of the powerful programming languages such as Python. This workshop is designed to introduce you to the world of data analysis and to show you how powerful Python is as a programming language for each segment of data analysis. We will start the lectures with topics about setting up the work environment (Jupyter Notebook & Python), after that we will move on to data preparation for data processing and data cleaning (pre-processing part). When the data is prepared for work, we will first analyze the data using descriptive data analysis and visualization techniques, then using inferential statistics methods (e.g. T-test, ANOVA, MANOVA, Mann-Whitney U etc.), and finally we will go through both correlation (Pearson correlation coefficient and Spearman’s rank correlation coefficient) and Linear regression analysis. After each of the conducted statistical analyses, we will present a report with the interpretation of the obtained results and their applications in business. ​In accordance with the progress of the participants, we will cover more advanced lessons (topics – like Time Series Analysis etc.) on data analytics in Python.
Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
09:00

 —

11:30

Analyze your data in Python

Isidora Gataric
Senior Data Analyst @ JAGGAER

Object detection using OpenCV

Image Classification, Image Processing, Machine Vision, Computer Vision
Library Demonstration
Beginner to Intermediate
Object detection model can identify which of a known set of objects might be present and provide information about their positions within the image
Image Classification, Image Processing, Machine Vision, Computer Vision
11:45

 —

13:15

Object detection using OpenCV

Amir Siddiqui
Data Scientist @ T Systems

Introducing two evaluation criteria for the next data prediction using Machine Learning models

Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
Tool Showcase
Intermediate
In the era of exponential growth of artificial intelligence, there is a need for an objective evaluation of the performance of machine learning models. It is a well-known fact that the performance of models directly depends on the database on which they were trained. In accordance with this, the same well-known databases are used in the scientific community so that the results of the validation of models of different authors are comparable. When publishing results with new databases or different data-preparing processes it is necessary to introduce new evaluation criteria that directly link the characteristics of the database and the obtained machine learning metrics. This lecture gives an example of how this can be done in the case of predicting the following data from a machine learning model.
Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
13:30

 —

15:00

Introducing two evaluation criteria for the next data prediction using Machine Learning models

Dusan Radivojevic
Machine Learning Engineer and System Admin @ Vinca Institute of Nuclear Science

Introduction to data mining (especially for healthcare domain)

Data Wrangling, Data Mining, Edge Computing
Tool Showcase
Beginner
The aim of the tech tutorial is overview and short introduction to data analysis and data mining for students and professionals out of area of computer science and data science. The tech tutorial is primarily formed for professional from healthcare. The showcase will be on dataset from the healthcare domain with aim to show significance of data analyses and such knowledge for establishment of diagnoses, finding therapies and medicines and much more… Milena would be happy to share her knowledge with public before she goes on PhD studies in Barcelona. The programming knowledge is not necessary because the showcase will be with graphical data mining tool Knime or Orange. The idea of this approach is to encourage professionals to sail into the amazing world of data science.
Data Wrangling, Data Mining, Edge Computing
15:00

 —

16:30

Introduction to data mining (especially for healthcare domain)

Milena Stojic
Teaching associate @ Faculty of Mathematics, University of Belgrade

Stream 2

Predicting Natural Hazards using satellite imagery and AI

Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
Career Path, Library Demonstration
Intermediate
The Himalayan region of Uttarakhand is prone to recurrent landslides causing extensive loss of lives and properties. The fragile Himalayan ecosystem because of its unstable slopes, untamed infrastructural development and intense monsoonal precipitation makes it highly susceptible to landslides. This tutorial attempts to integrate Geospatial Intelligence (Geospatial data along with Artificial Intelligence) to predicting regions of landslide prone areas in the Rishikesh – Gangotri highway to generate a landslide hazard zonation for safety and mitigation purpose. It makes use of scikit learns machine learning library specially ANN and XGBOOST along with geospatial modules such as GDAL and rasterio for handling the geospatial raster data, geopandas for handling the vector shapefiles and matplotlib for plotting the geospatial data .The tutorial will be beneficial in replicating other natural disasters like flood and earthquakes also.
Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
09:00

 —

11:30

Predicting Natural Hazards using satellite imagery and AI

Rahul Das
Scientific OffIcer @ Flood and River Erosion Management Agency Assam

Big data analysis with MapReduce

Data Wrangling, Data Mining, Edge Computing
Career Path, Library Demonstration
Intermediate
This tutorial provides an overview of MapReduce, a powerful paradigm widely used for processing large scale datasets. Attendees will gain understanding of the core principles of MapReduce, its benefits for big data processing and its practical application in problem solving. The tutorial will cover algorithms which use MapReduce, implemented in Python using popular frameworks like Apache Hadoop, Apache Spark and MRJob.
Data Wrangling, Data Mining, Edge Computing
11:45

 —

13:15

Big data analysis with MapReduce

Ana Vranic
Reasearch Assistant @ Institute of Physics, Belgrade

Simplifying Event-Driven Services with FastStream: A How-To Guide

Data streaming
Library Demonstration
Beginner
Join us in this comprehensive tutorial as we explore FastStream, a powerful Python framework that streamlines event-driven service development. FastStream encapsulates multiple brokers, including KafkaBroker (formerly FastKafka), offering versatility in messaging solutions. We’ll guide you through harnessing FastStream to simplify your event-driven services, covering essential concepts, implementation, and practical examples. Discover how to leverage FastStream’s capabilities to efficiently build event-driven services, eliminating the complexities of communication between services. We’ll delve into its elegant design, explore key components, and demonstrate its ease of use.
Data streaming
13:30

 —

15:00

Simplifying Event-Driven Services with FastStream: A How-To Guide

Tvrtko Sternak
Software Engineer @ airt technologies

Automation of the Automation using ChatGPT

Natural Language Processing, Conversational AI, Speech Recognition, Machine Translation
Library Demonstration, Tool Showcase
Beginner to Intermediate
Tips for prompt engineering; how to use chatGPT for generating autotests; integrating with existing tools and libraries such as Playwright; adding self-testing and self-healing mechanism; what’s next: using ChatGPT API and other genAI tools, adding to a real project, adding DSL for generating tests, covering both UI and API testing and etc.
Natural Language Processing, Conversational AI, Speech Recognition, Machine Translation
15:15

 —

16:45

Automation of the Automation using ChatGPT

Gennadii Chursov
Mentor @ h.careers

Stream 3

Build End-To-End MLOps Platform with Open Source DVC Ecosystem

Image Classification, Image Processing, Machine Vision, Computer Vision, Machine Learning, Deep Learning, GNNs, CNNs, RNNs, MLOps
Career Path, Library Demonstration, Tool Showcase
Intermediate to Advanced
“This talk discusses the landscape of features and uses cases of DVC in ML engineering and MLOps: – Get started with versioning data, artifacts, and models with DVC and DVCLive – Automated and reproducible pipelines with DVC – Experiment management and metrics tracking with DVCLive and VSCode – Setup CI/CD to test and deploy ML models with DVC and CML – Model performance and data drift monitoring with DVC The talk summarizes the experience of more than 2 years experience on integrating DVC in a number of ML projects around various industries: Banking, Telecommunication, Automotive, Retail, Health, Public Sector. Attendees will get practical insights into the features and limitations of the tool for different ML applications: computer vision, computer vision, NLP, batch scoring, etc. “
Image Classification, Image Processing, Machine Vision, Computer Vision, Machine Learning, Deep Learning, GNNs, CNNs, RNNs, MLOps
09:00

 —

11:30

Build End-To-End MLOps Platform with Open Source DVC Ecosystem

Mikhail Rozhkov
Solutions ML Engineer @ Iterative.ai

Some Implications for the use of Gen AI in courts

Ethics and Legal
Tool Showcase
Advanced
“The Fundamental Rights are the foundation of the constitutional system of Democratic States, and the Due Process, the backbone of the Judiciary’s performance. Discussion about the process is fundamental to improving justice, especially in a context of technological and social changes. The use of AI technologies in the judicial system, for example, can bring significant advances in terms of efficiency, but it also raises quesƟons about the protection of fundamental rights of citzens. Behold, AI technology can be used as a perpetrator or perpetrator of violations of Fundamental Rights, which makes it essential to discuss ways to regulate its use and guarantee the protection of rights; it is therefore as fallible as the person behind it. Cultural and legal differences between the countries involved can generate divergences in the approach to these issues, making it even more important to hold an international congress to promote the exchange of knowledge and experiences among participants. “
Ethics and Legal
11:45

 —

13:15

Some Implications for the use of Gen AI in courts

Thiago Felipe Avanci
Researcher, Professor, Lawyer @ Nastri & Avanci

Building Trustworthy and Ethical Generative AI for Enterprise Applications

Natural Language Processing, Conversational AI, Speech Recognition, Machine Translation
Tool Showcase
Beginner to Intermediate
In this session I will show attendees how they can implement generative AI into their business, based on their enterprise data. I will show how they can use IBM Watsonx.ai platform and use IBM LLM Granite, other third party LLMs or their own LLM for classification, summarization and content generation.

My colleagues Dora Doljanin and Marko Pap will help with the demo.
Natural Language Processing, Conversational AI, Speech Recognition, Machine Translation
13:30

 —

15:00

Building Trustworthy and Ethical Generative AI for Enterprise Applications

Jelena Skalec
IBM Data & AI Technical Sales @ IBM Croatia

Augmented Reality – ARKit framework showcase

Image Classification, Image Processing, Machine Vision, Computer Vision, Augmented reality
Tool Showcase
Intermediate
Application of augmented reality in modern applications from scratch using Apple’s ARKit library. We will dive deep into framework’s capabilities.
Image Classification, Image Processing, Machine Vision, Computer Vision, Augmented reality
15:15

 —

16:45

Augmented Reality – ARKit framework showcase

Marko Mutavdzic
Software Engineer @ FaceFiltersPro
Download

BREAKOUT SESSIONS

Stream 4

STREAM 5

STREAM 6

Breakout Break BS3

11:00

 —

11:30

Breakout Break BS3

Download

CO-LOCATED EVENTS

Stream 4

GenAI on Vertex AI: From zero to hero

Generative AI
Career Path, Library Demonstration, Tool Showcase
Intermediate
“Generative AI (GenAI) is a rapidly developing field that is revolutionizing the way you build innovative applications. In this hands-on session, you will learn how to use GenAI on Vertex AI to create your own generative AI applications. You will start by learning the basics of GenAI, including different types of GenAI models and how they work. Then, you will dive into hands-on tutorials where you will learn how to use GenAI on Vertex AI to generate text, code, images, audio and more. By the end of this session, you will have a solid understanding of GenAI and how to use it to create your own innovative applications.”
Generative AI
09:00

 —

11:30

GenAI on Vertex AI: From zero to hero

Ivan Nardini & Thu Ya Kyaw & Lavi Nigam
Customer Engineer @ Google Cloud & Senior Developer Relations Engineer @ Google Cloud & Machine Learning Engineer @ Google Cloud

Transaction log as the heart of Delta Lake

Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
Library Demonstration, Tool Showcase
Intermediate
The transaction log is the core mechanism that empowers Delta Lakes. The goal of the session is to examine the structure and the usage of the transaction log in Delta Lake and to briefly mention the improvements that Delta Lake brings over the standard data lakes. In this session we will use Apache Spark and (mainly) PySpark library to perform various operations on Delta Lake. After every execution, we will observe the changes in the transaction log and explain what they are used for and why they are important.
Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
11:45

 —

13:15

Transaction log as the heart of Delta Lake

Boris Matijasevic
Full Stack Software Engineer @ CodeEssence

Complex networks sparsification: methods and tools

Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
Tool Showcase
Beginner to Intermediate
One of the most prominent features of real complex networks is that they are sparse. However, we often lack direct information about the real connections between nodes, and we need to infer them from the data. Consequently, we obtain dense networks containing inessential or noise edges, such as complex networks of words or online social networks. In this tutorial, we will present some of the network sparsification methods and demonstrate their effectiveness in distinguishing between essential and non-essential edges.
Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data
13:30

 —

15:00

Complex networks sparsification: methods and tools

Marija Mitrovic Dankulov
Research Professor @ Institute of Physics Belgrade

Molecule Testing for Ansible-Driven Data Lake Deployments on Kubernetes

Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data, Testing
Tool Showcase
Intermediate
Dive into automated data lake deployments. This session integrates Molecule with Ansible, Vagrant, and VirtualBox for streamlined Kubernetes setup on a single virtual machine—ideal for testing. Learn to configure VirtualBox with Vagrant, leverage Molecule for Ansible testing, and swiftly deploy six node Kubernetes cluster. Explore essential services—Elasticsearch, Kibana, JupyterHub, Apache Spark, and Apache Airflow. Gain insights into setting up and optimizing data services for testing. By the end, adeptly incorporate Molecule testing into your Ansible workflow, ensuring a robust foundation for trials.
Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data, Testing
15:15

 —

16:45

Molecule Testing for Ansible-Driven Data Lake Deployments on Kubernetes

Jakob Triva
Consultant @ Poslovna Inteligencija d.o.o.

stream 5

Hands on Quantum Computer

Quantum Computing
Tool Showcase
Intermediate
Implementing quantum program from zero to production. In this tutorial, we will learn how to assemble a quantum circuit, encode classical data into a quantum state, connect to a quantum backend and execute the circuit, decode the results, and pack all of this into a single Python function that can be integrated into any application.
Quantum Computing
09:00

 —

11:30

Hands on Quantum Computer

Petar Korponaic
Founder & CTO @ Quantastica

Create ML framework – introduction to no-code ML

Natural Language Processing, Conversational AI, Speech Recognition, Machine Translation, Image Classification, Image Processing, Machine Vision, Computer Vision, Machine Learning, Deep Learning, GNNs, CNNs, RNNs
Tool Showcase
Beginner to Intermediate
Create ML is a framework by Apple allowing for simplified manipulation of the training process. Users can effortlessly build and train models with little to no previous knowledge. The Create ML framework is integrated into Xcode, and allows for seamless integration into applications built for Apple’s ecosystem. The framework provides models for image, video, motion, sound, text, and tabular based predictions. The aim of this workshop is to demonstrate these features and apply them to a Swift application built in Xcode. The workshop is indented for beginners in the field of AI, but can prove interesting to the more advanced audience as the no-code trend is on the rise.
Natural Language Processing, Conversational AI, Speech Recognition, Machine Translation, Image Classification, Image Processing, Machine Vision, Computer Vision, Machine Learning, Deep Learning, GNNs, CNNs, RNNs
11:45

 —

13:15

Create ML framework – introduction to no-code ML

Aleksandar Petrovic
Teaching Assistant @ Singidunum University

Building and Optimization of Neural Networks from Scratch with Numpy

Machine Learning, Deep Learning, GNNs, CNNs, RNNs
Career Path
Beginner to Intermediate
During this tutorial we will start from the basics, participants will learn how to create neural network architectures from scratch and implement forward and backward propagation. We will cover optimization techniques like gradient descent, weight initialization, and regularization. By the end, attendees will have a strong foundation in building and optimizing neural networks without relying on deep learning frameworks.
Machine Learning, Deep Learning, GNNs, CNNs, RNNs
13:30

 —

15:00

Building and Optimization of Neural Networks from Scratch with Numpy

Marina Marjanovic
Professor @ Singidunum University

Model validation using deep generation of stress data

Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data, Machine Learning, Deep Learning, GNNs, CNNs, RNNs
Tool Showcase
Intermediate
The tutorial is aimed to present a novel procedure for model robustness validation. The key idea is to apply deep generative models that can sample unlikely events and introduce slight shifts in input data. Such an approach allows us to examine models’ reactions to input data with different levels of the likelihood. The growing number of running models highlights the importance of model validation, which determines the process of verifying the validity of input and output data, model’s performance, stability, and interpretability.
Data & Predictive Analytics, Data Visualisation, Time Series, Geospatial data, Machine Learning, Deep Learning, GNNs, CNNs, RNNs
15:15

 —

16:45

Model validation using deep generation of stress data

Vitaliy Pozdnyakov
Junior Research Scientist @ AIRI

stream 6

Download
Billetterie Weezevent