How is the first conference different from the second? When preparing the first one, there is neither the ability to “do as before”, nor the audience feedback, and the organizers have to make bold assumptions. This does not mean that it will turn out badly. But this means that after the first time, new considerations will certainly appear that will help in the second.
Last year we held the SmartData conference for the first time, and now the time has come for the second time: we are announcing
SmartData 2018 , which will be held
on October 15 in St. Petersburg .
What to expect from the conference? Who should go on it? What has changed since last year? We tell everything under the cut.
the main thing
The format of the conference is best known to the Program Committee - people selecting reports and helping to improve them. Therefore, we asked the participants of the PC and supplemented this text with their direct speech.
Roman p0b0rchy Favorite , who during the preparation of the first SmartData gave us
an interview , was again the most talkative of all.
The first noticeable difference of the new SmartData is that the
site immediately shows the enumeration “for whom this conference is”: Data Scientist, Data Engineer, Data Architect and Statistician. A year ago, it was just a “data conference.” What is behind the addition of specific roles?
Roman: “Last time we received a lot of feedback from viewers, which is not clear for whom this conference is: for a data engineer or for a data scientist? The program included reports on both these areas, but it was not obvious to people.
Now we want to explicitly indicate that the event is for those and for others. And we are trying to do just that for him: so that within SmartData both a full-fledged conference for an engineer, and a full-fledged conference for a scientist will be obtained. So that in every time slot any of them had something to listen to. ”
The second thing that can be noticed on the site: if a year ago, under the name SmartData, the subtitle “On Big and Smart Data” flaunted, now it has been replaced by the inscription “Conference for those who are immersed in the world of machine learning, analysis and data processing”. What is behind this?
Tanya Denisyuk (Program Coordinator): “For the first time, we had positioning uncertainty, and in the description there was a“ big data ”phrase that could attract“ random ”people. But the conference is not about HYIP at all; it requires immersion in the world of working with data. Therefore, now we have removed this phrase, and in general have changed the positioning, so that tech-savvy people come. ”
And what other changes can we expect?
Roman: “First of all, last time all the reports were Russian-speaking, and now we want to bring well-known foreign speakers. If possible, even those who are not very often in Russia. It’s too early to promise anything, but work is underway.
And secondly, we want the conference to include not only reports, but also some other interesting “magic” associated with data processing. Perhaps in the form of interactive things that can be poke in the breaks. And perhaps in the form of discussions, where participants will be able to declare topics interesting to them. ”
Concrete
Since there are still a few months before the conference, most speakers and reports will be known closer to the point. But now we can give examples to get some idea about the program:
The analysis of the texts is an area of ambitious and important: it is obvious that humanity has not achieved much here, but progress is being made, and its achievements can be of great benefit to humanity. All this is interesting both from a scientific point of view and applied.
Such a subject requires a speaker of the appropriate caliber, and it will be on SmartData.
Konstantin Vorontsov is one of the most famous Russian specialists involved in machine learning. Moreover, he sees the picture at once from both sides, “academic” and “industrial”: Konstantin is both a senior researcher at the Computing Center of the Russian Academy of Sciences, and a Yandex expert.
Who can know about face recognition more than others? Someone whom users download tens of millions of photos daily.
Alexander Tobol , who works at Odnoklassniki, now knows how to effectively implement facial recognition, even with large scales and high loads. And on SmartData will share knowledge with others: what to use, how to optimize, when to train.
Those who were on the previous SmartData are not required to represent
Ivan Yamshchikov : his performance
“Why did we teach the neural network to write poems in the style of Kurt Cobain?” Led the audience rating. And for those who were not there, we add that Ivan works at the Max Planck Institute, and he is interested in the principles of the work of artificial intelligence, which could help to understand how our brain works.
Now he will give a very different report, and here’s how Ivan himself describes it: “We’ll talk about how the use of machine learning in the harsh world of an enterprise differs from B2C, we’ll figure out whether it’s possible to build AI solutions in the face of data shortages, and discuss the best the practice of using machine learning in “combat” conditions using examples of ABBYY products. ”
How can the Yandex logo appear in an Apple presentation? Thanks to the CatBoost gradient booster library: when we started talking about machine learning on the WWDC 2018 keynote, its name got on one slide with such giants as TensorFlow and Caffe.
Last year,
Anna Veronika Dorogush from the CatBoost team had already
talked about the project on SmartData, and then after the report she was completely surrounded by questions. Since then, the library has got both new features and new stars
on GitHub - apparently, it will be very useful again, that there are special discussion zones for speakers on SmartData.
And also we draw your attention to the fact that now the
reception of reports is in full swing. So, if you work with data and you have something to share, do not hold in yourself and contact!
Who are the judges?
This year there are many new faces in the SmartData program committee. And since it depends on these people what to expect from the program, it makes sense to understand from which side they themselves look at the data processing tasks. We asked the PC members to tell us briefly about themselves.
Anton ZlodeiBaal Maltsev : “I have been doing computer vision for 10 years now. We (two or three classmates from the Physicotechnical Institute) do either tasks for the order or our own models, modules, which we then sell and implement. Something in between a start-up and custom development. For 10 years, our solutions have already been implemented in many places. First of all, it is different biometrics associated with computer vision: recognition of the iris, veins of the hands, a bit of face recognition. There are also solutions with recognition of goods on the shelves, license plates, trains ”.
Alexander Sterligov : “I used to work in Yandex, I was in charge of managing clouds. This is due to the data in the sense that the clouds are a large number of servers, they break, there are problems, you need to deploy your software on each server, and this all somehow rests on the data. Now I work in a Joom project and am responsible for the entire analytical platform and infrastructure: data collection, data acquisition, processing. I provide tools for the work of analysts and data scientists. ”
Alexey Tikhonov : “I work in Yandex: for five years I was in search analytics, now in design analytics. In addition, it is affiliated with such projects as Yandex.Autopoet, and crafts in the spirit of "Neural Defense" and Neurona.
Pavel Plotnikov : “I work at Wrike, I started as an automator in the analytics department, then I became a data engineer. I spent a lot of time in an embrace with Spark, then tried to make a Mesos cluster to plan Spark tasks, distributing resources for them. And at the same time they made a platform that helps launch services (like the one in Kubernetes, it was not so popular at that time). And now I’m called analytics ops, because with the data engineers divided into several sub-departments, and now I’m more involved in infrastructure and tools. ”
Nikita Povarov : “I at JetBrains are committed to trying to replace machine learning heuristics in IDE (where it makes sense). Previously, in Yandex, I was engaged in metrics in A / B experiments, in general A / B experiments, and now sometimes I consult some external companies on this topic. ”
Roman Poborchy : “I also worked in Yandex for many years, all the time I was there with people who did all sorts of interesting things related to data, and I myself was involved in it. It also seems to me that I learned, in the first approximation, to distinguish the present from the unreal, and I want to organize a conference, where only the present will be, without a hyip, for those people who are really doing something. It was partially possible for the first time, but now I want to do it completely. ”
And
Vitaly Khudobakhshov remained unanswered, but you can already know him by the
habrapost and the
report about how the probability of being in a relationship depends on the person’s name. These posts with a report appeared thanks to the work of Vitaly on Odnoklassniki, and now he has a new main activity: he is engaged in the development of Data Science-tools in JetBrains.
Tickets
With tickets compared to the previous year, there are also changes.
That's what remains the same: as the conference approaches, their price grows, so it makes sense to buy as early as possible (the nearest price increase is already July 1).
But what is changing: now, instead of a single price for all, several options have appeared. For participants who work in a large company and go to the conference at its expense, in general, everything is the same. For small businesses that are more aware of the cost of tickets, now there is a discount. And for those people who go "on their own", the discount is even greater.
Tickets are now on sale
on the site . And all the additions to the program will also appear there. So, if you have already decided to go to the conference, follow the link now, and if you need more data for this, open it in the future.
See you on SmartData!
