The workshops from the list below are the complimentary part of full tickets.
The recordings of most workshops will be shared with full ticket holders after the conference.
Workshops will be run remotely via Zoom.
Introduction to Machine Learning on the Cloud
This workshop will be both a gentle introduction to Machine Learning, and a practical exercise of using the cloud to train simple and not-so-simple machine learning models. We will start with using Automatic ML to train the model to predict survival on Titanic, and then move to more complex machine learning tasks such as hyperparameter optimization and scheduling series of experiments on the compute cluster. Finally, I will show how Azure Machine Learning can be used to generate artificial paintings using Generative Adversarial Networks, and how to train language question-answering model on COVID papers to answer COVID-related questions.
- - Introduction to AI in the Microsoft Azure Cloud
- - Introduction to Azure Machine Learning
- - Training Titanic predictive model using Automatic ML
- - Using Jupyter Notebooks in Azure ML
- - Using Visual Studio Code to schedule and track experiments
- - Hyperparameter optimization and HyperDrive
- - Training GANs on Azure ML to produce Artificial Art
- - Using DeepPavlov NLP to answer COVID-related questions on CORD dataset
To actively take part in this workshop, you will need Microsoft Azure Account. In order not to spend time during the workshop, please register free Azure account in advance if you do not already have one (students can register student account).
Beginners and ML practitioners.
Nov 2, 18:00-20:00 CET. The workshop will be recorded.
AI Assistant with DeepPavlov
Simple chatbots can be built with a number of simple rules, but their UX will be very narrow and limited. Building complex AI assistants, in contrast, requires a lot of effort. To make your users happy, you have to master many things at once: rules, intents, finite state machines, chit-chat tech, dialog flow management, and many other things. It's a job for an entire team. Fortunately, you don't have to hire it. Instead, you can rely upon your trusty software from DeepPavlov.ai. With it, you can offload the complexity of managing the entire infrastructure to DeepPavlov Agent – an open-source Conversational AI orchestrator. Next, you can pick state-of-the-art NLP components from DeepPavlov Library to decode human emotions and understand their intents. Finally, you can pick some of the available skills from DeepPavlov Dream to avoid reinventing the wheel like chit-chat.
In this workshop, you'll be guided through the entire process, from an overview of the platform to building a sample AI assistant that you'll be able to re-use later in your own work.
By the end of this workshop you'll be able to use the provided components from the DeepPavlov family of projects to build simple AI assistant:
- work with existing NLP components from DeepPavlov Library like sentiment and emotion classifiers
- build simple goal-oriented skill with DeepPavlov Go-Bot
- combine them together with the DeepPavlov Agent into a simple AI assistant
- augment your simple AI assistant with some of the DeepPavlov Dream pre-built skills
Intermediate or Advanced.
Nov 3, 18:00-21:00 CET. The workshop will be recorded.
From Shazam to Today's Audio Fingerprinting Algorithms, the Theory, Best Practices, and Open Source Libraries to Experiment with
An acoustic fingerprint is a condensed digital summary, a fingerprint, deterministically generated from an audio signal, that can be used to identify an audio sample or quickly locate similar items in an audio database. Using open-source python libraries, we will describe the process of creating an audio fingerprint from a media file, how to store it, and search for matches.
- - 1st part. An introduction to audio fingerprinting–one of the first algorithms developed in the industry and the most successful commercially until today was developed by researchers from Shazam. Their solution is to identify the strongest peaks in the spectrogram and to store the relative signatures of these peaks.
- - 2nd part. Shazam allows you to send a recording made on your phone of nearly any song–not covers, and it will tell you the song’s name, and other data about that song. We will describe the process of creating an audio fingerprint from a media file, how to store it, and search for matches.
Nov 4, 17:00-18:30 CET. The workshop will be recorded.
Hands on with TensorFlow.js
Come check out our workshop which will walk you through 3 common journeys when using TensorFlow.js. We will start with demonstrating how to use one of our pre-made models - super easy to use JS classes to get you working with ML fast. We will then look into how to retrain one of these models in minutes using in browser transfer learning via Teachable Machine and how that can be then used on your own custom website, and finally end with a hello world of writing your own model code from scratch to make a simple linear regression to predict fictional house prices based on their square footage.
Workshop requires some working knowledge of using HTML / CSS / JS to get the most out of it. No Machine Learning background is assumed.
Great for beginners or researchers looking to get the reach and scale of web technologies so more people can use their models. No ML knowledge is assumed.
Nov 4, 18:00-21:00 CET. The workshop will be recorded.
Applied AI with AutoML and Interpretability in MATLAB
Are you new to Machine and Deep Learning and want to learn how to apply these techniques in your work? Building good machine learning models is difficult and time consuming – get empowered to apply AI to your domain with AutoML and tools that don’t require extensive coding experience!
- - Part I (first hour): Seminar on how Automated Machine Learning (AutoML) can help practitioners and engineers use AI in their domains even though they lack experience with machine learning or are concerned about the black box nature of AI models. AutoML simplifies model development to a few steps, and interpretability can draw back the curtain from complex black box models.
- - Part II (2 hours): Hands-on workshop where you work in our MATLAB Online platform and build machine learning models using interactive tools. You will apply machine learning to the problem of classifying human activities based on accelerometer sensors, get familiar with the basics of deep learning, and apply transfer learning to an image processing task.
No machine learning background is assumed. No prior experience with MATLAB is expected, though casual knowledge of a high-level language will be helpful.
To actively take part and access MATLAB Online, you will need a MathWorks account. If you do not already have one, please register by following this link.
Great for beginners who need to familiarize themselves with the basics, but also researchers and engineers eager to learn about how advanced methods like AutoML and model interpretability will help them apply AI to their domains.
Nov 9, 18:00-21:00 CET. The workshop will be recorded.
The Hitchhiker's Guide to the Machine Learning Engineering Galaxy
Are you a Software Engineer who got tasked to deploy a machine learning or deep learning model for the first time in your life? Are you wondering what steps to take and how AI-powered software is different from traditional software? Then it is the right workshop to attend.
The internet offers thousands of articles and free of charge courses, showing how it is easy to train and deploy a simple AI model. At the same time in reality it is difficult to integrate a real model into the current infrastructure, debug, test, deploy, and monitor it properly. In this workshop, I will guide you through this process by sharing tips, tricks, and favorite open source tools that will make your life much easier. So, at the end of the workshop, you will know where to start your deployment journey, what tools to use, and what questions to ask.
- - AI-powered software versus traditional software
- - MLOPS and AI pipeline in a nutshell
- - ML Platform: what is this?
- - Hands-on with AI pipelines
To get the most out of the workshop working knowledge of Python is required. No Machine Learning background is assumed.
Great for software engineers who got tasked with AI model deployment for the first time. No ML knowledge is assumed.
Nov 10, 18:00-20:00 CET. The workshop will be recorded.
Crowdsourcing Practice for Efficient Data Labeling
In this tutorial, we present a portion of unique industry experience in efficient data labeling via crowdsourcing shared by both leading researchers and engineers from Yandex.
We will make an introduction to data labeling via public crowdsourcing marketplaces and will present the key components of efficient label collection. This will be followed by a practice session, where participants will choose one of the real label collection tasks, experiment with selecting settings for the labeling process, and launch their label collection project on one of the largest crowdsourcing marketplaces. The projects will be run on real crowds within the tutorial session. While the crowd performers are annotating the project set up by the attendees, we will present the major theoretical results in efficient aggregation, incremental relabeling, and dynamic pricing. We will also discuss the strengths and weaknesses as well as applicability to real-world tasks, summarizing our five year-long research and industrial expertise in crowdsourcing. Finally, participants will receive a feedback about their projects and practical advice on how to make them more efficient.
Beginners, advanced specialists, and researchers are invited to learn how to collect high quality labeled data and do it efficiently.
Nov 11, 18:00-21:00 CET. The workshop will be recorded.