Online practical conference about ML, AI and Deep Learning applications

Machine Learning Prague

, 2021

Buy your ticket now 50% off

We can't fix 2020, but a better ML Prague? That we can do!

ML Prague will run online to assure your attendance is 100% safe. This even allows us to bring you even more practical content! At the same time, this will be the most interactive ML Prague ever, including deeper discussions with our speakers after each talk, mastermind sessions, networking activities with peer-experts from the whole world, and a hackathon before the conference. Stay tuned for more information on what's coming!

Note: If you registered for ML Prague 2020, your ticket is still valid for our online conference on February 26-28, 2021. You'll find your gift below, under our conference program section.

  • 1000 Attendees
  • 3 Days
  • 45 Speakers
  • 10 Workshops
  • 1 Hackathon

Phenomenal Speakers

Practical & Inspiring Program


Room 103 Room 106 Room 203 Room 205 Room 206

Agile Data Annotation

Room 103

Marek "Marx" Grac, Phalanx

Come join us for our workshop and get hands-on experience with data annotation. The main goal of data annotation in Machine Learning algorithms is to make the implicit explicit so that the learning process can be improved. Even though many people see data annotation as a mundane task the process of creating guidelines and processes can be very interesting. In this workshop you will test various data annotation techniques mainly application-driven and low-cost approaches. We will also focus on how to measure the quality of the resulting data as well as test various UX principles and see how much they impact the cost-efficiency. Finally when you get bored of doing the manual part of data annotation yourself we will go through the basic legal aspects of outsourcing it.  

Automatic and Explainable Machine Learning with H2O

Room 106

Jo-fai (Joe) Chow, H2O

General Data Protection Regulation (GDPR) is now in place. Are you ready to explain your models? This is a hands-on tutorial for beginners. I will demonstrate the use of open-source H2O platform ( with both Python and R for automatic and interpretable machine learning. Participants will be able to follow and build regression and classification models quickly with H2O's AutoML. They will then be able to explain the model outcomes with various methods.

Machine Learning in Julia

Room 203

Kevin O'Brien, Coillte
Avik Sengupta, Julia Computing

Julia is specifically designed from the start of its conception as a language for high-performance computation but at the same time highly interactive. To achieve this Julia is one of the few modern languages that relies in just-in-time (JIT) compilation via LLVM to make its code run as fast or faster than statically compiled C and fortran codes. Its modern language design has the following features: multiple dispatch Lisp-like macros dynamic types type inference built-in parallel/distributed computing lightweight threads and elegant high-level language constructs. Outline: Introduction to Julia The Julia Language Julia in Data Science Julia Interfacing with Python and R Machine Learning in Julia High-performance computing in Julia  

Data Analysis in Big Data Environment with Apache Spark and Python

Room 205

David Vrba, Socialbakers
Peter Vasko, Socialbakers
Jiri Harazim, Databricks

Apache Spark became a standard for data processing and machine learning in big data environments and is popular especially for its high-level DataFrame API that allows working nicely with structured data in a very efficient way. In the first part of this workshop we will get familiar with the DataFrame API of Spark and see some challenges that you might face when processing large datasets. We will explore some advanced optimization techniques and see how to apply them to compose efficient analytical queries. In the second part of the workshop we will see how Spark can be used for machine learning and deep learning in particular. We will explore Deep Learning Pipelines - a library that integrates Spark with deep learning frameworks such as TensorFlow and Keras.

Programming the Pepper Robot

Room 206

Aleš Horák, Informatics at Masaryk University
Adam Rambousek, Faculty of Informatics at Masaryk University

The social robot by Softbank Robotics denoted as Pepper will be introduced. The robot hardware capabilities as well as examples of natural human-machine interaction in English and Czech (which are being developed by the team at FI MU) will be presented in detail including a tutorial on your own programming for a virtual or a real Pepper robot. The 1.2-m-tall robot is designed for social interactions with people and it is equipped with an extensive API set to detect faces mood or age and to react to their values.


Zero to AI: Workshop on the Wolfram Language

Room 103

Mark Braithwaite, Wolfram

Designed by Wolfram data science experts this workshop will provide an introduction to machine learning techniques illustrated with live dynamic examples using the Wolfram Language. The workshop will walk you step-by-step through the basics of machine learning methodologies and techniques and how to apply them using the Wolfram Language. Upon completion you will come away with enough practical knowledge to immediately use the Wolfram Language for your own machine learning tasks on text data or images including supervised classification and prediction unsupervised feature identification sequence prediction and computer vision.

Developing Autonomous Vehicles with High Fidelity Simulation

Room 106

Ashish Kapoor, Microsoft

High-fidelity simulations can provide a rich platform to develop autonomy by enabling the use of AI technologies such as deep learning computer vision reinforcement learning etc. We have developed AirSim which is a simulator for autonomous vehicles built on the Unreal Engine. It is open-source cross platform and supports hardware-in-loop simulation thus allowing rapid development and testing of the system. The simulation is developed as a plugin and can be simply be dropped into any Unreal environment. AirSim supports AI development capabilities by exposing APIs to enable data logging and controlling vehicles in a platform independent manner. We will give an overview of how to use AirSim for building realistic simulation environments and doing development for quadrotors that use popular flight controllers such as Pixhawk. It is developed as a plugin that can simply be dropped in to any Unreal environment you want. We will also showcase how the system can be used to incorporate machine learning components useful for building such autonomous systems.

Cloud-native AI on OpenShift

Room 203

Václav Pavlín, Red Hat
Francesco Murdaca, Red Hat

Ever thought of doing a cloud-native AI work? What does that even mean? This workshop will introduce you to running AI related services like Spark Seldon or Jupyter on Kubernetes as part of a project Open Data Hub. You will learn how to move your AI workloads to the cluster and implement a basic data science workflow. As Jupyter notebooks have become the de facto standard in data science we will show you how to use them and adopt some of the best practices that we’ve developed over time.

How to Make Data-Driven Decisions: The Case for Contextual Multi-armed Bandits

Room 205

Michal Pleva,, O2 Czech Republic
Petr Stanislav,, O2 Czech Republic

Supervised learning has done wonders but it’s fundamentally limited. A good prediction of customers' churn or the likelihood of new acquisition may not always help you to do what is best in a given situation. By attending our workshop you will get hands-on experience with algorithms for direct optimization of decision-making with uncertainty. We will be focusing on the special case of reinforcement learning known as Contextual multi-armed bandit problems. Those problems arise frequently in important industrial applications played a role in AlphaGo success and are very often adopted by industry leaders such as Google and Netflix. Decision making with uncertainty is a challenge so we will show you how to effectively balance between trying new things to find better solutions and repeating the behavior that works well. During the workshop you will have an opportunity to play with a linear algorithm to solve a simple problem as well as with more advanced solution involving a deep neural network to learn a latent representational feature space for a problem.

SAS Viya & Open Source Integration focus on Python

Room 206

Ivan Kasanicky, SAS
Jordan Bakerman, SAS

In this course you will learn to use the Python API to take control of SAS Cloud Analytic Services (CAS) actions. You will also learn to upload data into the in-memory distributed environment analyze data and create predictive models in CAS using familiar Python functionality via the SWAT (SAS Wrapper for Analytics Transfer) package. You will then learn to download results to the client and use native Python syntax to compare models.


Welcome to ML Prague

AutoML with Keras Ecosystem

Haifeng Jin, Google

The Keras ecosystem now has two new members, Keras Tuner and AutoKeras. They are built with AutoML techniques to dramatically reduce the manual work for designing and training deep learning models. They work seamlessly with Keras and TensorFlow for model export, saving, and deployment. The talk not only covers how to use them but their underlying mechanism as well.

Autonomous driving: few insights on perception and explainability

Matthieu Cord, Valeo

Self-driving is a safety-critical application. In this talk, I first present the machine learning framework used for autonomous driving, gathering contributions from computer vision, deep learning, and autonomous robotics research fields. I then discuss some of the main challenges we face at to improve advanced driver-assistance systems. I will give some examples such as unsupervised domain adaptation for visual segmentation, or driving behavior explanation system using natural language processing.

Private Federated Learning

Vojta Jína, Apple

Federated Learning is a new approach that is picking up steam in the machine learning community as a way to improve global models by leveraging on-device training on user data. At WWDC 2019, Apple announced Private Federated Learning by combining Federated Learning with Differential Privacy. We have started to use this technology in iOS 13 for a variety of use cases, including QuickType keyboard, Found in Apps, and Smart Replies. In this talk, Vojta will provide more details about this technique.


Marketa Hulik Jansova, Tibor Zavadil: Automation of quality assessment of data from solar stations
Chandra Mohan Meena, Sarwesh Suman: Clustering Maturity of Machine Learning Platform Users
Rafał Bachorz, Małgorzata Mochol-Grzelak, Grzegorz Miebs: Efficient strategies of static features incorporation into the Recurrent Neural Network
Arun Mathew: SAP Behavioral Insights
Dominik Krzemiński: U-Net for Automated Segmentation of Knee Cartilage Imaging
Adarsh Prasad Behera, Shekhar Verma, Manish Kumar: Manifold Visualization using Discrete Surface Ricci Flow
Andreea Mihailescu: Designing Predictive and Recommender Systems for Optimized Laser-Plasma Interaction Experiments. Implementations, Challenges and Future Prospects.
Stephan Baier: Analyzing Customer Feedback for Product Fit Prediction
Réda Alami: Memory bandits for decision making in non-stationary environment

Ensuring Machine Learning Fairness with Monotonic Constraints

Serg Masís, Syngenta

The first part of the session underpins the importance of Machine Learning Interpretation. Fundamentally, it is needed because machine learning by itself is incomplete as a solution. After all, the problems they solve are not deterministic, so the solution cannot cover all of it because it is an optimization. One of the most significant issues is that AI faces today is overconfidence. Given the high accuracy of AI solutions, we tend to increase our confidence level to the point we fully understand the problem. Then, we are misled into thinking our solution covers all of it! The machine learning interpretability toolkit can help us first learn from our models.  Then, leverage what was learned or our domain knowledge to place guardrails, mitigate bias, and enhance model reliability, making them safe to use even in rare and unexpected situations and free from non-discriminatory practices. One of the ways in which fairness can be ensured is through monotonic constraints. We will discuss several scenarios in which this may be needed. 

During the second part, we will dive into a law school scholarship problem. Let’s suppose a law school wants to handout merit-based scholarships to those students most likely to pass the bar exam. To that end, they want to train classifiers that score students on this probability. However, the classifier must be consistent with merit-based norms such as having the highest grades in other examinations. Employing monotonic constraints in XGBoost and Tensorflow will place the guardrails so that students with high examination scores are never unfairly penalized.  We will walk through the code that assesses, establishes, and confirms model fairness.

How to build the perfect model of a human according to their voice

Petr Schwarz, Phonexia

Voice biometry is a technology that overperforms humans. Petr Schwarz will present how modern voice biometry systems are built and how they are deployed. The key issues are how to collect data, what are the input features describing the human vocal tract, what machine learning techniques are used for modeling, how to train the models, and how to deliver the model to its user while keeping the best accuracy.

Expertise recommendations - A supervised approach that surmounts incomplete datasets

Jeremy Jonas, McKinsey & Company
Felipe Vianna, McKinsey & Company

For knowledge-based organizations, finding precise expertise to address specific projects is increasingly important. At McKinsey we’ve been improving our internal expertise search capability, by enriching colleague profiles in various ways, including ML-driven recommendations for ‘Topics to call me about’. Through a number of innovations, our Prague-based Data Science team has created a highly-effective prediction model.
Traditionally, expert profiling and retrieval are based on document retrieval approaches. But can the information available in profiles be used to train a supervised model? As with many retrieval applications, our challenges began with a limited amount of data available, as well as the format, which at McKinsey is mostly PowerPoint files. Several well-known approaches were combined to perform a Document Classification step in unsupervised fashion, providing data to create the expert-candidate representations. In a later step, profiling of the experts was achieved despite noisy label data (incomplete profiles) and a large amount of features (compared to the amount of samples available).
Despite these challenges, the model is now achieving 80% acceptance of recommendations. In turn this is materially helping the Firm find appropriate experts when needed.


The state of production machine learning in 2020

Alejandro Saucedo, Institute for Ethical AI & Machine Learning

As the number of production machine learning use-cases increase, we find ourselves facing new and bigger challenges where more is at stake. Because of this, it's critical to identify the key areas to focus our efforts, so we can ensure our machine learning pipelines are reliable and scalable. In this talk we dive into key trends for production machine learning in 2020. These include machine learning deployment, interpretability, orchestration and concept drift.

Continuous Machine Learning

Paweł Redzyński,

In the software engineering world, CI/CD practices have proven to be a reliable and effective approach to automating recurring tasks, like running tests, code analysis checks and even delivering final products to production. In this talk, we will present how to automate ML processes using GitHub Actions or GitLab CI/CD and Continuous Machine Learning (CML) library that will take care of:
• transferring large datasets to CI runners
• managing GPU/CPU resources for computations
• generating ML model report with metrics and plots right in GitHub Pull Request
so that ML specialists can focus on research.

Exchange rate estimation of cryptocurrencies with FPGA and GPU accelerators

Krzysztof Rojek, byteLAKE

Predicting cryptocurrency exchange rates is a significant issue in the economic market. A cryptocurrency is a digital asset designed to work as a medium of exchange that uses strong cryptography to secure financial transactions, control the creation of additional units, and verify the transfer of assets. In current world, it gives serious effect on worldwide economic aspects. For the last years we observe an intensive development of accelerated computing platforms. Although current trends indicate a well-established position of GPU devices in the HPC environment, FPGA aspires to be an alternative solution to offload the CPU computation. This work presents a machine learning-based approach to predict the exchange rate of current cryptocurrencies in order to develop a better understanding on its price. Our goal is to compare the possibilities of utilization of platforms equipped with the FPGA and GPU accelerators. We focus on the set of aspects taking into account adaptation of ML methods to the accelerators, performance of the proposed solutions, accuracy of results, as well as the energy consumption of the platforms. This work investigates the most advanced accelerators used in the ML solutions, including NVIDIA Tesla V100 GPU, and Xilinx Alveo U250 FPGA. The exchange rate estimation is based on ML models using Long Short-Term Memory, as well as Neural Network with Hidden Layers.


The Ethical aspects of Machine Learning

Uri Eliabayev, Machine and Deep Learning Israel

Machine Learning has become a major part of our lives. As more and more companies and organizations implementing ML-based solutions, we need better understand the ethical aspects of Machine Learning algorithms.

In this talk, we will speak about the key element of this field (Fairness, explainability, bias and more) and give some past examples of ethical problems in the ML field. Alongside that, We will suggest ways to solve or reduce the ethical problem in each ML project and finally, we will learn how companies like Google and Microsoft make their algorithms fairer.

Building Augmented Analytics for 50% of Fortune 100

Filip Dousek, Workday

One year ago, Prague-based were acquired by Workday. Today, the same team is building its ML-driven augmented analytics for Workday's largest customers. Filip will talk about the concept behind, how it is different and why it's called the next generation of BI&analytics.

Conference day 1

Building Safety Mechanisms in Autonomous Systems

Ashish Kapoor, Microsoft

Machine Learning is one of the key component that enables systems that operate under uncertainty. For example, AI systems and robots might employ sensors together with a machine learned system to identify obstacles. However, such data driven system are far from perfect and can result in failure cases that can jeopardize safety. In this talk we will explore a framework that aims to preserve safety invariants despite the uncertainties in the environment arising due to incomplete information. We will describe various methods to reason about safe plans and control strategies despite perceiving the world through noisy sensors and machine learning systems. We will also consider extensions of these ideas, using high-fidelity simulation, to a sequential decision making framework that considers the trade-off in risk and reward in a near-optimal manner.

Understanding and mitigating unwanted bias in Artificial Intelligence

Karthikeyan Natesan Ramamurthy, IBM Research AI

AI and machine learning models are increasingly used to inform high-stakes decisions. Discrimination by AI becomes objectionable when it places certain privileged groups at systematic advantage and certain unprivileged groups at systematic disadvantage. In this talk, we will discuss the sources of unwanted bias in AI, and how it manifests along various points in the AI pipeline. We will also explore several methods of bias mitigation. Finally, we will discuss how bias can be measured and mitigated using the open source AI Fairness 360 toolkit.

Harnessing relational learning for explainable learning

Tomas Pevny, Avast

While most of the machine learning methods assume that samples are vectors, matrices, or sequences, in many real-world problems they have a rich structure. While this structure makes the manual design of features non-trivial, I see it as an inductive bias that should drive the design of models. In this talk, I will introduce a simple, yet powerful framework for learning on structured data. A side, yet important feature is the explainability of decisions, which is the result of ingesting data as-is instead of devising artificial features. A concrete implementation of the framework will be demoed on data from various stages of analysis of malware.


Modular MLOps architecture built to last

Radovan Parrák, Credo

Every company that takes machine learning seriously needs to ‘productionalize’ their ML pipelines. Efficiently, robustly and at scale. The emerging methodology called Machine Learning Operations (MLOps) comes to rescue.

However, there are already hundreds of convenient, stand-alone and overlapping ML tools, workflow managers, automation and orchestration frameworks, developed by both vendors and the open-source community striving to put this methodology into practice. New ones keep on appearing (and disappearing). As a result, many companies are contemplating whether to buy an MLOps platform or to build one internally. And if the latter then they hope to postpone the architectural decisions until the sheer amount of available options reduces to a widely accepted set of tooling. But will it ever?

In this talk, I will share some of Credo’s experience on how to design a modular and future-proof MLOps platform, based on open-source tooling, that hits the ground running today and survives still tomorrow in the everchanging zoo of ML tooling.

ML powered Crime Prediction

Or Herman-Saffar, Dell

What if we could predict when and where the next crimes will be committed? Crimes in Chicago is a publicly published dataset which reflects the reported incidents of crime that occurred in Chicago since 2001. Using this data, we would like not only to be able to explore specific crimes to find interesting trends, but also predict how many crimes will be taking place next week, and even next month.

Complex Systems for AI

Tomas Mikolov, Facebook AI Research

Machine learning has been tremendously successful in the last decade. The core concepts for training the models is to use supervision, error backpropagation and stochastic gradient descent. However, many scientists believe that to make steps towards more autonomous AI systems, we need to discover learning approaches that are fundamentally less supervised than the current ones. In this talk, I will describe a project where we attempted to define a system which can evolve for indefinitely long, possibly reach arbitrary complexity, and use no supervision. It is based on an old idea of cellular automaton which can be seen as a special type of a recurrent-convolutional network. I will show examples of interesting behavior that we did observe in automatically constructed models. We were able to discover these interesting automata using a novel metric which measures structured complexity growth in time. This work could be a basis of a new generation of machine learning models which can continue learning in interesting ways in situations where no supervision of even rewards are available.


Martin Holeček: Table understanding in structured documents
Jakub Slovan, Jan Rus, Luboš Andert and Petr Jančařík: Bayesian Social Media Content Inspiration
Christoph Briese, Erik Irmscher: Robust detection of non-cooperative aircraft using deep convolutional neural networks
Sebastian Eresheim: Cybersecurity Containment Agent
Gergely Szerovay, Laszlo Siller: An incluse open-source component library that bridges the gap in the application of machine learning tools
Martin Plajner: Generic system for promotional sales prediction from time series data and individual observations.
Jakub Bartel, Matej Choma, Vojtěch Rybář, Petr Šimánek: ML for High-Resolution Rainfall Forecast
Nóra Sándor, David Hagenmüller, Guido Pupillo: A machine-learning approach to mesoscopic Physics

Deep Neural Networks Abstract Like Humans

Hava Siegelmann, University of Massachusetts Amherst

Deep neural networks (DNNs) have revolutionized AI due to their remarkable performance in pattern recognition, comprising of both memorizing complex training sets and demonstrating intelligence by generalizing to previously unseen data (test sets). The high generalization performance in DNNs has been explained by several mathematical tools, including optimization, information theory, and resilience analysis. In humans, it is the ability to abstract concepts from examples that facilitates generalization; this presentation describes DNN generalization from that perspective. A recent computational neuroscience study revealed a correlation between abstraction and particular neural firing patterns. We express these brain patterns in a closed-form mathematical expression, termed the “Cognitive Neural Activation metric” (CNA) and apply it to DNNs. Our findings reveal parallels in the mechanism underlying abstraction in DNNs and those in the human brain. Beyond simply measuring similarity to human abstraction, the CNA is able to predict and rate how well a DNN will perform on test sets, and determines the better  network architectures for a given task in a manner not possible with extant tools. These results were validated on a broad range of datasets (including ImageNet and random labeled datasets) and neural architectures.

AIOPS, Machine Learning and Anomaly detection, our experience implementing a virtual assistant engine to detect and triage anomalous behavior in a data center

Kirill Maiantsev, Broadcom

Join this session to learn about our experience and challenges in designing a virtual assistant engine used in many of our IT data center monitoring products. We will discuss both the use case we have solved, its evolution as we encountered challenges and some of the machine learning models implemented to solve these problems. We will also provide in-depth review of the Kernel Density Estimation model we used to study time series in order to obtain the expected value ranges. When anomalies are detected in the expected values, we perform some additional learning on these anomalies to detect patterns and auto-correlate live events in the data center. During the session, we will share both our learnings and challenges building this enterprise grade virtual assistant.

Confidence Estimation Learning for Production-ready Neural Networks

Adam Blažek, Iterait

Deploying your ML model to production may bring you headaches for many reasons, e.g. out-of-distribution input data or low-quality user input. Recognizing those cases is a crucial step for providing actionable feedback and handling those cases properly. This talk uncovers our simple yet effective recipe for integrated confidence estimation learning alongside a practical example used in a production environment.


Panel Discussion

Ashish Kapoor, Microsoft
Karthikeyan Natesan Ramamurthy, IBM Research AI
Tomas Mikolov, Facebook AI Research
Hava Siegelmann, University of Massachusetts Amherst

Closing Remarks

Have a great time A great gift for this year’s attendees

Did you get your ticket before November 25, 2020? Then you’ll get a 50% discount to purchase your ticket for ML Prague 2022 to celebrate the return to the house of Machine Learning in CE, the Rudolfinum Music Hall!

Now or never Tickets

Standard Ticket

Late Ticket

Last 300 tickets

  • Conference days € 150
  • Only workshops € 195
  • Conference + workshops € 295

What You Get

  • Practical and advanced level talks led by top experts
  • Connect with ML pros from all around the world to share expertise
  • Access to actionable practical content

They’re among us We are in The ML Revolution age

Machines can learn. Incredibly fast. Faster than you. They are getting smarter and smarter every single day, changing the world we’re living in, our business and our life. The artificial intelligence revolution is here. Come, learn and make this threat your biggest advantage.

Our Attendees What they say about ML Prague

Thank you to Our Partners

Platinum Partner

Gold Partners

Silver Partners

Media Partners

Communities and Further support

Become a partner

Happy to help Contact

If you have any questions about Machine Learning Prague, please e-mail us at


Jiří Materna
Scientific program & Co-Founder

Gonzalo V. Fernández
(Communities & Media partnerships)

Teresa Caklova
(Event production)

Jona Azizaj