On November 8 and 9, a conference was held in Moscow with leading Russian and international experts in the field of AI and data analysis, as well as representatives of leading companies in the development and application of AI technologies in business processes. We share with you video recordings of presentations and presentations of our colleagues.
The first day of the conference was devoted to technical presentations, and the second to the application of technology in business. Technical presentations were divided into sections “recommendation systems”, “natural language processing”, “computer vision”, “new methods of machine learning”, etc. And business presentations were devoted to the peculiarities of the application of AI technologies in various industries.
Convergence of technologies as a trend in the development of artificial intelligence
Vladimir Novoselov
Vladimir, drawing on the experience of creating complex predictive analytics products for business in the Predict Mail.ru Group, said that one of the key trends in the application of machine-based services is their layout and mutual integration with other technologies and solutions, and not use as "Things in oneself." Only then, machine learning begins to bring maximum benefits for a real business, and allows you to become a real AI-native company.
Recording from the start of the performance
Presentation
Pulse recommendation system trends overview
Andrey Murashev
Andrey made an overview of such trends: multi-armed bandits, explanatory recommendations, privacy, fairness recommendations, optimization of several tasks at the same time, work with biased combat data and neural networks. Andrew described the strengths and weaknesses of these approaches, where they are most often applied.
Recording from the start of the performance
Presentation
Neuromachine translation in question-answer systems
Fedor Fedorenko
Fedor’s presentation was devoted to a review of problems and tasks in the field of machine translation, which were solved or are being solved today by the developers of several question-answer systems created in the Mail.ru Group. First, Fedor described the essence of the question-answer system, gave a simplified scheme of work, and then went through a number of practical tasks related to processing requests and issuing answers.
Recording from the start of the performance
Presentation
The world through the eyes of neural networks
Danila Baygushev and Alexander Snorkin
In the field of machine learning, there are problems when we need to know not only the forecasts of models, but also how they came to such decisions. This allows us to improve model algorithms. Danila and Alexander talked about how this can be done, and illustrated with the example of their working projects.
Recording from the start of the performance
Presentation
Ethics of Artificial Intelligence
Alexander Karmaev
Alexander spoke in part with philosophical topics: what is artificial intelligence and where is it needed? What interests should be considered by companies engaged in the development of artificial intelligence? Who should determine the framework in which artificial intelligence develops? Are the concepts of ethics applicable and which ones? Is ethical decision-making artificial intelligence for a person?
Recording from the start of the performance
Presentation
Our colleagues will speak at other AI Journey series events on AI and data analysis:
Yekaterinburg, November 28
Kuzma Khrabrov (research programmer)
“Generative-competitive networks. Theory and practice."
St. Petersburg, December 7
Mikhail Maryufich (developer OK)
"OCR 5 plus"
Yuri Makarov (OK developer)
“Musical recommendations in ok.ru: how we increased the quality at times, withstanding the load”
More about the project - ai-journey.ru/about-project