Today, machine learning is one of the fastest growing areas of computer science, in terms of both theoretical research and practical applications. Based on this trend, numerous companies are eager to get involved by organizing machine learning competitions for the students at Poznan University of Technology.

Roq.ad, Europe’s leading cross-device technology provider, and Poznan University of Technology are currently working in tandem to organize and bring students the “Let’s Roq” challenge. The competition’s key task for students is to find the best supervised algorithm for cross-device connections. It’s a very relevant and challenging real-life problem we face with today due to the rapid growth of the online advertising industry.

As a company, Roq.ad is a successful start-up headquartered in Berlin. Its engineering team is, however, almost entirely based in Poznan. That being said, the solutions proposed by Roq.ad are also strongly based on machine learning technologies.

During competition, students will face the task of estimating the probability that two given devices belong to the same user. It is worth noting that data made available for students is coming from actual Internet traffic. Of course, all sensitive information has been removed to prevent identification of devices and users.

All students of Poznan University of Technology are invited to participate in the competition (including graduate students). The student teams will compete for a prize pool of 10 000 PLN and gain the opportunity to complete an internship within Roq.ad’s data science team.

For more information about the competition, please consult the following sources:
https://challenge.roq.ad and
https://www.cs.put.poznan.pl/kjasinska/lectures/mmds-challenge/letsroq/

Competition Launch:

November 4, 2016, at. 9:45, room 125BT (during the lecture: The Mining of Massive Datasets Challenge taught by dr. Krzysztof Dembczyński).

Consultation hours (to discuss problems and get valuable hints or tips):

Tuesdays, 16:50-18:20 or 18:30-20:00, room 43 CW (supervised by Kalina Jasinska, M.Sc.)