If we recall the practice of data analysis 10 years ago and compare it with what is now, it becomes obvious — during this time, Data Science has come a long way. Computer vision, recommender systems, big data, artificial intelligence - in 2010 these words were used mainly only by leading IT companies. No one could imagine that in just 10 years, these technologies will change the world.
What would Netflix be without a recommendation system? Who will tell what series to watch in the evenings. Or Apple music, in which you are not told anything about the new Christian black metal style albums? Just think how long will it take to issue a loan without using a scoring system? Imagine YouTube showing nothing in the Featured Videos section. Although ... in this scenario, I would have slept more and not watched funny vidos about cats until 3 in the morning. The world expects unmanned vehicles to replace drivers, although in 2010 it was science fiction. Why is it that Tinder picks up pairs based on machine learning algorithms, people get married, they have children, if you think about it, it turns out that in fact they are children of artificial intelligence Sic.
We owe a lot to Data Science, so on December 16 we will get together at Odnoklassniki’s Moscow office and together with colleagues from OK, Sberbank, VK and X5 Retail Group we will hold ok.tech: Data Talk # 4 New Year's release. Let's talk about the results of the year and decade in the field of data analysis and processing. What was the industry before, what is it now, and what surprises await us in the future when Elon Musk replaces people with huge humanoid robots. You will receive answers to all these questions at our event.
Come! It will be useful, interesting and fun!
Register for the event.
Under the cutscene you will find descriptions of reports and a schedule.
CV: yesterday, today, tomorrow - Arthur Kuzin, X5 Retail Group
Over the past ten years, computer vision has given the world a wealth of new results. Algorithms bypassed a person in accuracy so many times, firmly entrenched in new areas and learned how to create new content.
The report will provide an overview of the evolution of neural networks in the problems of classification, detection, segmentation, video processing. The brightest results with generative models will be shown. As well as a small overview of the hardware on which the algorithms are running.
Data Science 2010-2019: a productive decade - Dmitry Bugaychenko
At the turn of the decade, we will try to look back and recall how the date of the Sainz evolved from Netflix prize to AlphaZero. Using OK.ru examples, let us see how the algorithms and approaches, as well as the generation of the Scientists' data, have changed. Let’s try to highlight the most significant events and see how they changed the views of datasaentists and society on life.
The history of the search for the ideal ML target - Andrey Yakushev, VK
In the report, we will tell you about the stages of growing the ML target on which Smart Tape is built - the most popular section of VKontakte. How we started with the sorting of posts by probability to put “I Like”, how we came to the statement of the problem in terms of multi-criteria optimization and how we solve it. And also try to answer the question - is it possible to automate the process of searching for ML target and what is needed for this.
Report Schedule:
19:30 - 19:40 Opening remarks, Alexey Chernobrov
19:40 - 20:00 CV: yesterday, today, tomorrow - Arthur Kuzin, X5 Retail Group
20:00 - 20:20 Data Science 2010-2019: a productive decade - Dmitry Bugaychenko
20:20 - 20:40 Search history for the ideal ML target, Andrey Yakushev, VK
20:40 - 21:00 Coffee break, networking
21:00 - 22:00 Discussion about the past, present and future Data Science
We are waiting for everyone who wants to talk with experts in the field of data analysis on December 16 at their Moscow office.
Come, it will be interesting and useful!
Register on ok.tech: Data Explained # 4 New Year's Edition
Event partners:
AI / Big Data / Machine Learning Channel
Educational portal GeekBrains