Three weeks ago, we finished the second edition of GHOST Day: Applied Machine Learning Conference and it’s time for a little summary. Probably the largest Polish AI published by the OPI National Research Institute in Warsaw, entitled the article about our conference “World-class AI in Poznań” and we think that this is a very accurate description of the conference program we constructed. The list of speakers from all around the world (USA, UK, France, Australia, Switzerland) from top R&D institutions such as Facebook AI, DeepMind, IBM Research, Google or New York University resulted in high satisfaction of the participants. The NPS satisfaction index of conference was 70 (a value above 50 should be considered as “excellent”). In addition, 100% of surveyed participants want to visit us again at a potential future edition next year. The organization of the conference was rated at 4.95 on a 5-point scale, and there were no major technical problems during the conference.

The conference was granted the honorary patronage of the Ministry of Economic Development, the Mayor of Poznań and the Polish Association of Artificial Intelligence. In addition, 13 ML organizations from all over Poland agreed to be a partner of the event, which allowed to reach the assumed number of participants, and the last tickets were distributed through selective registration. Ultimately, 389 participants attended the conference.

Apart from the organization, GHOST memebers also presented at the conference their incredible projects! Zuzanna Trafas gave a presentation on the methods of explainable artificial intelligence, which is one of the priorities of AI development according to EU. An extremely interesting presentation was given by Max Adamski on the use of Bayesian methods in the analysis of energy data, and Piotr Wyrwiński presented machine learning methods for detecting malware. The student session was very popular – over 100 people watched the presentations! Moreover, Adam Wojciechowski presented a poster about a new method of imbalanced image classification. Congratulations to all of you!