Practical conference about ML, AI and Deep Learning applications

Machine Learning Prague 2019

, 2019

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.

  • 1 000+ Attendees
  • 3 Days
  • 45 Speakers
  • 8 Workshops
  • 2 Parties

Phenomenal Speakers

Practical & Inspiring Program

Friday,
Workshops

at CEVRO Institut, Jungmannova 28/17, Prague 1

Registration

Room 103 Room 106 Room 203 Room 205 Room 206
coffee break

Machine Learning in the real-world: why you need a platform to run Machine Learning at scale (1/2) (sold out)

Greg Antell, BigML
Jose Antonio Ortega Ruiz, BigML

When data-driven organizations start using Machine Learning at scale (i.e. automatically creating and operating machine-learned models for dozens of different use cases) they soon realize how important it is to leverage a robustly-engineered platform that removes the complexities of Machine Learning and allows different stakeholders to focus on what matters most; enhancing and automating decision making.  In this workshop we will give you an extensive overview of the BigML platform and discuss the design principles that have been followed to remove the extra complexities that developing end-to-end Machine Learning applications typically implies while ensuring traceability reproducibility and many other features that mitigate the risks of operating hundreds thousands or even millions of machine-learned models in the real world. 

Quantum Enhanced Optimization and Machine Learning (1/2) (sold out)

Jacob Biamonte, Skolkovo Institute of Science and Technology

This course introduces contemporary methods to utilize quantum processors to accelerate computational subtasks. No background in quantum mechanics is assumed. As part of a unified quantum programming framework we will introduce the core building blocks needed to program two types of quantum devices. (Type I) Quantum enhanced annealers; (Type II) gate-model based devices. We will compare and contrast the use of quantum annealers and gate-model based devices to (i) solve optimization problem instances; (ii) train binary classifiers. Course notes problem sets with solutions as well as compete programming examples will be provided.

Development of AI sound analysis method for automotive aftersales care (sold out)

Jiří Dobeš, GoodAI Applied
Jan Šinkora, GoodAI
Martin Poliak, GoodAI

Follow our journey from a good accuracy ML method to a very fine-tuned product with hands-on workshop.

Deep Learning for Text Processing (sold out)

Petr Baudiš, Rossum
Martin Holeček, Rossum

Join us for our workshop and get hands-on experience with Neural Network models used in Natural Language Processing (NLP). We will introduce you to the most common Neural Nets and techniques used in NLP (bag-of-words embeddings CNNs RNNs and autoencoders). Aside of basic text classification (similar to last year workshop) we will also talk about measuring text similarity as another task. We will be playing with some of the most popular tools like Keras and Tensorflow.

Machine Learning on the EDGE (sold out)

Jan Pospíšil, Microsoft

Come to see how you can deploy and host ML model(s) on edge Raspberry Pi like devices. I will walk you thru full end-to-end deployment process using Azure IoT infrastructure. To “save” some time we will be using ML models from Custom Vision yet this scenario is definitely not restricted to those models only.   Workshop Environment: Visual Studio Code https://code.visualstudio.com/ Edge device (Raspberry Pi) or Linux virtual machine (Can be provisioned locally or on Azure as same as on any public cloud). Raspberry Pi should be running Raspbian https://www.raspberrypi.org/downloads/raspbian/ (Lite version recommended) VM should be running Ubuntu (16.04 or 18.04) https://www.ubuntu.com/#download Ideally have Azure environment (Free version is more than OK) https://azure.microsoft.com/en-us/https://www.raspberrypi.org/downloads/raspbian/ (Lite version recommended) VM should be running Ubuntu (16.04 or 18.04) https://www.ubuntu.com/#download Ideally have Azure environment (Free version is more than OK) https://azure.microsoft.com

Lunch
coffee break

Machine Learning in the real-world: why you need a platform to run Machine Learning at scale (2/2) (sold out)

Greg Antell, BigML
Jose Antonio Ortega Ruiz, BigML

When data-driven organizations start using Machine Learning at scale (i.e. automatically creating and operating machine-learned models for dozens of different use cases) they soon realize how important it is to leverage a robustly-engineered platform that removes the complexities of Machine Learning and allows different stakeholders to focus on what matters most; enhancing and automating decision making.  In this workshop we will give you an extensive overview of the BigML platform and discuss the design principles that have been followed to remove the extra complexities that developing end-to-end Machine Learning applications typically implies while ensuring traceability reproducibility and many other features that mitigate the risks of operating hundreds thousands or even millions of machine-learned models in the real world. 

Quantum Enhanced Optimization and Machine Learning (2/2) (sold out)

Jacob Biamonte, Skolkovo Institute of Science and Technology

This course introduces contemporary methods to utilize quantum processors to accelerate computational subtasks. No background in quantum mechanics is assumed. As part of a unified quantum programming framework we will introduce the core building blocks needed to program two types of quantum devices. (Type I) Quantum enhanced annealers; (Type II) gate-model based devices. We will compare and contrast the use of quantum annealers and gate-model based devices to (i) solve optimization problem instances; (ii) train binary classifiers. Course notes problem sets with solutions as well as compete programming examples will be provided.

Distributed Learning and You (sold out)

Ruksi Laine, Valohai
Aarni Koskela, Valohai

We'll use common distributed learning libraries to speed up a machine learning training examples focusing on data parallelism and scalability of the training. After the workshop you will have the knowledge of why and how distributed learning can be used. Requirements: Laptop with Docker installed a dash of software development expertise and an open mind.

Analyzing social media data with Apache Spark using Python (sold out)

David Vrba, Socialbakers

Apache Spark is popular computational engine used for big data processing machine learning and streaming. In this workshop we will take a deeper look how Spark can be used for analyzing social media data and building machine learning models. First we will spend some time with the DataFrame API and use it to answer some analytical questions about the data. Then we will explore and run some graph processing algorithms using Spark library GraphFrames that is built on top of DataFrames. In the next part we will focus on ML Pipelines - Spark native library for machine learning we will explore it's basic concepts such as Transformer Estimator and Pipeline and use it to build some ML models on the data.

Artificial intelligence and ethics (sold out)

Ondřej Veselý, Kiwi.com
Tomáš Kopeček, Red Hat

In the practical part of the workshop we'll learn how EU currently evaluates the ethical dimension of ML related R&D projects and how European AI Alliance prepares the roots for eventual AI regulations. In the light of medialised ML related ethical cases we'll discuss how it relates to our job and how we can avoid unethical decisions. Following part focuses on the ethics itself from different philosophical-sociological aspects. We'll identify and go through four basic narratives of AI discourse: scientific religious securitization and artistic.

from
Party
La Loca Music Bar, Odborů 278/4, Prague 2

Saturday,
Workshops

Rudolfinum, Alšovo nábřeží 12, Prague 1

Registration

Welcome to ML Prague 2019

Why inverse reinforcement learning is impossible, and why we can do it anyway

Stuart Armstrong, Future of Humanity Institute

Most versions of IRL (inverse reinforcement learning) have assumed the human expert is rational, noisily rational, or, at best, fails to be rational is specific known ways. If the expert's rationality is completely unknown, however, then nothing can be said about their reward. More worryingly, simplicity priors don't help here - if we assume that the simplest interpretations are best, we get degenerate interpretations of human behaviour.
However, if we make a few a priori assumptions that cannot be derived from observations, we can do IRL successfully. One such assumption is that the mental models in human brains are informative about human preferences. This talk will illustrate how this can happen, and it's surprising connection with the "symbol grounding problem" - though applied to humans this time.

Data-driven System health determination in Monitoring Softwares for Operational Intelligence

Vítězslav Vít Vlček, Broadcom

The taxonomy of the different business problems from a data scientific perspective is relatively finite and it all starts with an accurate system health detection. However, the traditional methods of system health detection have practical limitations when monitoring performances of applications, infrastructure or network.  These limitations arise from the imposition of frequentist statistics based frameworks on common problems such as anomaly and outlier detection. In this session, we talk about some novel patented frameworks derived from quantum mechanics along with reinforcement learning algorithms to transition from the widely used frequentists statistics approach for system health detection to bayesian statistical approach. We also elucidate how at Broadcom we capture such frameworks to address multiple problems across lifecycle of monitoring softwares like anomaly detection, root cause analysis, predictions and prescriptions. 

Topological Approaches for Unsupervised Learning

Leland McInnes, Tutte Institute for Mathematics and Computing

Many topics in unsupervised learning can be viewed as dealing with the relative geometry of data. In mathematics, topology and homotopy theory are the fields that deal with similar kinds of questions. Using ideas, techniques, and language from topology can prove fruitful for unsupervised learning. This talk will look at how topological approaches can be brought to bear upon unsupervised learning problems as diverse as dimension reduction, clustering, anomaly detection, word embedding, and metric learning. Through the lens and language of topology and category theory we can draw common threads through all these topics, pointing the way toward new approaches to these problems. I hope to convince you that topological approaches offer a rich and growing field of research for unsupervised learning.

POSTER SESSION & LUNCH

Aliya Amirzhanova: Smart system for forecasting energy power consumption

 

Justin Shenk: Spectral decomposition for live guidance of neural network architecture design

 

Jan Rus: Social Media Influencers Recommendation Engine

 

Mladen Fernežir: Selling Second-Hand Items Faster by Using Deep Learning

 

Anastasia Lebedeva, Lenka Vraná: Processing medical records: turning archives into actionable insights

 

Myungsu Chae: Effective Network Compression Using Simulation-Guided Iterative Pruning

 

Jonas Bialopetravičius: Deep learning based star cluster parameter inference

 

Filip Široký: Anomaly Detection Using Deep Sparse Autoencoders for the CMS Data


Petr Sojka: Semantically Coherent Vector Space Representations

Parameter Servers Suck, All Hail Horovod

Ruksi Laine, Valohai

Ruksi will discuss using our supreme overlord **data parallelism** to shard training data to achieve tenfold machine learning experimentation iteration speed. We'll go through the newest trends in distributed training, with some examples and highlighting common bottlenecks.

Spot the villain – The Merlon Identity Index

Dušan Fedorčák, CEAi

At Merlon Intelligence, one of the CEAi portfolio startups, we are helping banks to fight money laundering by providing a system for screening of their potential customers. Currently, we process more than 10 TB of news articles and our main goal is to index the data for efficient search and risk identification. Although the process seems to be straightforward, there are many challenging problems hidden inside application of machine learning methods, in data annotation strategies or in performance and explainability requirements. I am going to talk about several of these issues and share our solution(s) from the first prototype to the production-ready system used in a very rigid banking environment.

Machine learning: Explainablity with anti-models

Srivatsan Santhanam, SAP

Assisted or Supervised models of machine learning are a great boon to automate many mundane tasks in Enterprise world. Supervised model combined with hypothesis models helps in creating “situations” and in predicting possible outcomes to a given situation, thereby making “situations” intelligent.
But the question in user’s mind is how can I believe/trust what’s proposed is correct?
It’s here explainability becomes critical in ML. Inherently Decision Trees, Random Forest and features like boosting does help in showcasing significant factors and does trace the path of decision. LIME (Local Interpretable Model-Agnostic Explanations) does play a good role here. But for a user that’s just 50%. The remaining 50% is why the other “possibly close match” was not selected. It’s here the concept of anti-model is used.
The anti-model addresses the question of why the other “possibly close match” was not selected say in a classification scenario by building anti-models to the original models and its inference models to “explain” the decision taken. Multiple anti-models are constructed and validated with original model.
We will look at this ML approach, its benefits, applicability and potential success SAP has obtained in this space.
 

COFFEE BREAK

Solving the Text Labeling challenge with EnsembleLDA and Active Learning

Alexander Loosley, Data Reply

Want to build a text classification pipeline and have text with high quality labels that business can act on?  Great, throw in a language model, some BiLSTMs and CNNs and viola, you have trained a high-quality classifier. 

Unfortunately, many text data available for industry projects are unlabeled and difficult to label because of their industry specific nature.  The challenge can be split into three parts:

1. Unsupervised Text Exploration – What types of texts are there?
2. Label Curation – Given the texts, which set of labels provides the most business value?
3. Active Labeling/Learning – Which texts should be labeled first/next when human labeling is expensive?

This talk shares a few technical stories for solving all three challenges.

The Labels are Out There

Lotem Peled, NLP Lecturer

Supervised Learning is all about the labels. From straightforward classification tasks to complex sequence to sequence models - we need labels in order to learn. Nowadays, some tasks have an abundance of labeled data available online, and for other tasks we usually just go for crowdsourcing. But what happens if both of these options are irrelevant? In this talk we will discuss exactly such cases, and present some creative ideas on how you can go along obtaining labeled data for your task.

Automated Dropbox files naming with image recognition

Vasily Korf, Datalore

JetBrains provides a comprehensive set of software development tools. Not so long ago, we wondered if we could develop a useful tool for our design team, too. Our designers’ technology stack is based mostly on MacOS, Adobe products, and Dropbox as the cloud storage. Designers are very talented and creative people; sometimes they might store their files with an inconsistent path or under an inconsistent name. After a while, finding some old file (among tens of thousands of others) when needed started to turn into a puzzle.

We came up with the idea to develop some deep-learning-based tool that could sort out all these files (raster and vector images along with PDFs, PSDs, and AIs!) and get some interesting stats about them.

COFFEE BREAK

Deep Neural Networks for Optical, Multispectral and Radar Satellite Imagery. Can GANs help us?

Jan Zikeš, SpaceKnow

In this talk we’ll demonstrate how SpaceKnow exploits the state-of-the-art deep learning techniques and the latest cloud technologies to generate insights and provide a global geospatial intelligence. We will focus mainly on applications related to the combination of optical satellite imagery, multispectral imagery and SAR. We will elaborate on how can recent models like GANs possibly help us.

Vehicles detection in aerial images leveraging advanced image preprocessing

Ondřej Székely, IBM

Computer vision in aerial imagery domain can be quite challenging for some specific use cases, specifically in public safety. Based on one of the client’s needs we needed to focus not only on the neural network architecture, but also image preprocessing which is sometimes omitted. We leveraged particular series of convolutional filtering and wavelets transformation which is space depending. 

In the presentation, we will go through the process of data collection. Show benefits of preprocessing techniques to increase visual information. Discuss wavelet transformation and why wavelet transformation was used. Demonstrate the app which was used to display results. 

Party

Techtle Mechtle, Vinohradská 47, Prague 2

Sunday,
Conference day 1

Rudolfinum, Alšovo nábřeží 12, Prague 1

Doors open at

Solving the 3 main theoretical puzzles of Deep Learning

Tomaso Poggio, MIT

In recent years, artificial intelligence researchers have built impressive systems. Two of my former postdocs — Demis Hassabis and Amnon Shashua — are behind two main recent success stories of AI: AlphaGo and Mobileye, based on two key algorithms, both originally suggested by discoveries in neuroscience: deep learning and reinforcement learning. There is, however, little in terms of a theory explaining why deep networks work so well. In this talk, I will review an emerging body of theoretical results on deep learning including the conditions under which it can be exponentially better than shallow learning. The class of deep convolutional networks represent an important special case that avoids the curse of dimensionality for the class of hierarchical locally compositional functions.

How to win Kaggle competition and get familiar with machine learning?

Marcin Szeliga, MS Consulting

Have you heard that knowing machine learning is the easiest way to get rich quickly? Let’s test this statement. Kaggle is the place to do data science projects, why not to start there? During this session we will solve simple Kaggle competition. Actually, we will submit two solutions. The first made with super-duper deep neural network (black-box approach). Then we will follow proven ML methodologies and solve the problem methodically. All that using SQL Server Machine Learning Services. Minimum slides and maximum fun guaranteed.

COFFEE BREAK

AutoML in Predictive Modeling

Pavel Kordik, RECOMBEE

Automated Machine Learning was gaining momentum in 2018 with several open source projects released from top AI labs. You will learn internals of AutoML algorithms, their advantages and limitations. We will focus mainly on predictive modeling, but also demonstrate our AutoML research in the field of data clustering and recommender systems.

Machine Learning for recommender systems

Marc Romeyn, Spotify

This talk will focus on all the engineering aspects involved in Machine Learning at scale. A common warning shared with aspiring Data Scientists & ML engineers is that 90% of the work is about gathering, cleaning and validating data plus deploying and monitoring models. Yet for a long time most of the open source ML tooling focused on the modelling part.

We will first give an overview of the different ML Engineering frameworks out there, both open and closed source. We will then focus in on Kubeflow Pipelines and TFX (Tensorflow Extended), both of which are open source and model agnostic, by giving an end-to-end example highlighting why these frameworks are incredibly powerful.

Throughout the talk we will work to implement an end-to-end example, a deep neural network for predicting customer satisfaction in taxi trips using a public dataset with over 100M rows from the city ofChigago. This example includes transforming and validating the data, training a model in a distributed way, validating and monitoring model performance and last but not least deploying the model."

Spelling Correction for Web Search

Vladimír Kadlec, Seznam.cz

Web search engine almost never searches (only) for the words that users type in. Instead, during query processing phase, the engine tries to predict additional terms to search for.  This is especially important for misspelled queries.  The talk will provide details about how machine learning can be used to "read" user's mind for a correct version of the query. It is not a neuroimaging presentation but rather a brief history and current research of the spelling error correction system in the Seznam.cz's search engine".

POSTER SESSION & LUNCH

Seema Chouhan: Contextual Deep Learning Framework for Joint Object Detection & Scene Classification of Ground Level Photos


Adrián Pallas: A Deep Learning approach to online monitoring laser-based 3D printing


Roi Méndez Rial: Real-time arc welding process monitoring using Deep Learning on an IoT OPC UA device


Lukas Sekanina: Towards Energy Efficient Deep Learning on a Chip

 

Fiona Dick: Identifying risk-factors: classification modeling based on single nucleotide variants from the Alzheimer's Disease Sequencing Project


Michal Štefánik: Video699: Interconnecting Lecture Recordings with Study Materials


Vít Novotný: Soft Cosine Measure: Capturing Term Similarity in the Bag of Words VSM

   
Hanna Kujawska: Learning rank aggregation methods

 

Karl Trela, Yuri Campbell: Using text-classification and recommender systems to find industry-partners for research organizations

ML powered Crime Detection

Kateřina Veselovská, Deloitte Central Europe

Robust early case assessment and search term identification has been a critical competency within forensic investigations for a long time. However, the volume of analyzed text data grows rapidly and a multi-language litigations are getting more and more common. Traditional approaches are getting costly and less practical and it is getting difficult to find a “smoking gun”. To discover patterns and trends in the data, it is no longer possible to rely on manual analysis only. This session introduces the state-of-the-art ML methods for automated unstructured content analysis within electronic discovery and crime detection as such.

Predicting the Global Economy with Automated Machine Learning for Dynamic Systems

Maksim Sipos, CAUSALENS

The world is becoming more dynamic and real-time. We see more and more services becoming on-demand including food, transportation, delivery etc. This trend will continue to grow and become dominant. 
For example, there are more than 5 million drivers working for ride sharing companies spending 20 millions hours on the road every day. In order to allocate them more efficiently, ride sharing companies need vast amounts of accurate predictive models that can adopt to a changing and dynamic world for each one of the cities they operate in.
We will explain how AUTONOMOUS predictive technology, utilizing Automated Machine Learning for Dynamic Systems, can solve the problem at a more granular level than ever before. The way to solve it is by automating the whole pipeline of the construction of predictive models, from data preparation to feature extraction, feature set selection, hyperparameter tuning and performance evaluation of the models in Real-Time.

Artistic applications of artificial intelligence

Luba Elliott,

Over the past couple of years, there has been increasing interest in applying the latest advances in machine learning to creative projects in art, music, theatre and beyond. From Google's DeepDream and style transfer to a GAN-generated painting selling for 400,000 USD at auction, more and more creative AI projects are moving beyond the world of research and academia into the public eye. This talk will give an overview of how artists and creative technologists are using machine learning and investigating its creative potential.

COFFEE BREAK

Panel Discussion

Tomaso Poggio, MIT
Srivatsan Santhanam, SAP
Stuart Armstrong, Future of Humanity Institute
Lotem Peled, NLP Lecturer

Closing Remarks

Have a great time Prague, the city that never sleeps

A unique capital where you can breathe centuries of history at every corner. We’ll take a tour to explore the sights, invite you to taste the best pivos (that’s beer in Czech) and bring you back to the present by clubbing with you the whole night!

Impressive Venue

Now the seat of the Czech Philharmonic, the Rudolfinum is a Neo-renaissance building associated with music and art since 1885 and only used for truly outstanding purposes. The comfort, acoustics, and design makes it the greatest venue in the whole of Europe and it’s available for us.

Conference Hall

Rudolfinum
Alšovo nábřeží 12, Prague 1

Workshops

CEVRO Institut
Jungmannova 28/17, Prague 1

Now or never Tickets

Subscribe to our newsletter and we’ll let you know when tickets go on sale.

Early Bird

Sold Out

  • Conference days € 200
  • Only workshops (sold out) € 150
  • Conference + workshops (sold out) € 330

Standard Ticket

Sold Out

  • Conference days € 240
  • Only workshops (sold out) € 170
  • Conference + workshops (sold out) € 390

Late Ticket

Sold out

  • Conference days € 280
  • Only workshops (sold out) € 195
  • Conference + workshops (sold out) € 450

What You Get

  • Practical and advanced level talks led by top experts
  • 2 parties in the city with people from around the world. Let’s go wild!
  • Traditional Czech food throughout the conference

We Know That A Little Party Never Killed Anybody

Friday party 19:00

La Loca Music Bar
Odborů 278/4, Prague 2

Saturday party 19:00

Techtle Mechtle
Vinohradská 47, Prague 2

Our Attendees What they say about ML Prague

Thank you to Our Partners

Platinum Partner

Gold Partners

Silver Partners

Party Partner

Further Support

Become a partner

Happy to help Contact

If you have any questions about Machine Learning Prague, please e-mail us at
info@mlprague.com

Organizers

Šárka Štrossová
sarka@mlprague.com

Jiří Materna
jiri@mlprague.com