Report from Data Fest⁴ February 11-12

On February 11-12, the fourth Data Fest⁴ conference was held in our Moscow office, bringing together researchers, engineers and developers associated with Data Science in all its manifestations. Under the cut, we have prepared for you videos from the conference.

Speeches February 11

"ML-competition." What do machine learning contests give and what will they never learn at the university?

- “What does kaggle teach?”
Stas Semenov, Top-1 Kaggle


')
- “Machine Learning Competitions: Opportunities, Limitations, Pitfalls”
Alexander Fonarev, Rubbles SBDA Group

"ML in the industry." How does machine learning solve real practical problems of companies?

- “Technologies of hyper local weather forecasts”
Dmitry Solomentsev, Yandex



- “Segmentation of users in space-money dimensions”
Evgeny Nekrasov and Dmitry Bolkunov, Mail.Ru Group



- “Machine Learning in engineering and industrial applications”
Evgeny Burnaev, Skoltech, IITP

"ML in social networks." From the first person: what does machine learning look like in the two most popular social networks in Russia?

- “Machine Learning at VK”
Pavel Kalaydin, Vkontakte



- “Machine Learning at OK”
Dmitry Bugaychenko, Mail.Ru Group

"NLP". Recent achievements and results of new approaches to machine learning in the tasks of working with texts.

- “Deep Architectures for Natural Language Processing”
Sergey Nikolenko, HSE


- “Machine Learning behind Google Translate”
Mike Schuster, Google Brain

"NLP-tools." Surveys and master classes on modern methods of text analysis and processing and practical problem solving in NLP.

- “Vector representations of words and documents”
Anna Potapenko, HSE



- “BigARTM workshop”
Alexander Romanenko, MIPT

"Scraping and data collection." What tools can I extract?

- "How to collect dataset from the Internet in 2 parts"
Mikhail Korobov, Konstantin Lopukhin, Scrapinghub



- "Writing spiders, or what to do when you are calculated by IP"
Dmitry Sergeev, Zeptolab

Program for February 12

Data & Science. How to use machine learning in various scientific fields?

- “Scientific challenges to data analysis”
Oleg Bartunov, GAISH-MSU



- “Learning machine learning at the tasks of the Large Hadron Collider”
Andrey Ustyuzhanin, Yandex

"Make ML great again!" Analysis of famous myths, prejudices and bad practices, swept on the wave of popularity of DS, ML, BigData and AI. A series of 5 short speeches:

Vyacheslav Baranov, MyMind



Yuri Kashnitsky, Mail.Ru Group



Mikhail Trofimov, ML Works



Pavel Nesterov, freelance

"Artistic Intelligence". Reviews and demonstrations of the latest research and development in the machine transformation of media content.

- “Neural style transfer for music”
Dmitry Ulyanov, Yandex, Skoltech



- “Neural networks and creativity. Who is to blame and where is the line? ”
Ivan Yamshchikov, Yandex

"ML in science." In the continuation of the Data & Science section, we will examine case studies, tools and solutions with machine learning in science.

- “Tasks of Neuroscience, Neurohacaton Analysis”
Alexey Osadchy, HSE
(no video)

- “ML vs oncology: the tasks of design and optimization of therapeutic proteins”
Elena Ericheva, BrainGarden



- “The Path from Economics to Data Science”
Leonid Danilchenko, Game Insight

"ML Failconf". Speakers disassemble cases when machine learning turned out to be in the role of a microscope, which nails were hammered.

- "How to collect dataset from the Internet in 2 parts"
Ilya Saitanov, DSM Group



- "RL and shy market maker"
Andrey Chertok, Sbertech



- “Fedor score vs Machine Learning”
Dmitry Bugaychenko, Mail.Ru Group



- “Data Analysis on the Command Line”
Nikolay Markov, Aligned Research