Avast at Machine Learning Prague 2017

Avast technology is part of the next-gen artificial intelligence revolution.

Machines can learn incredibly fast. Faster than us, actually. They are already changing our world, our business, and our life. Avast, as a part of the artificial intelligence revolution, is proud to be a gold partner of Machine Learning Prague 2017, held from April 21st to April 23rd in Prague, Czech Republic.

Avast Threat Labs experts Martin Balek, Adam Hanka, Monika Seidlova, Ondrej Svec, and Tomas Trnka will be at the conference talking about machine learning in malware detection, specifically approximately the paradigm of zero FPR.

“This paradigm means that we want to be as sure as possible that we do not produce false detections,” explains Balek.

The one-daylight workshop offers a conservatory see at various algorithms used in malware detection and provides participants when hands-re experience in designing such algorithms based more or less a real-world dataset extracted from subsequent to more 200,000 malware and benign software samples. The practical part of the conference will be organized as a hackathon-following competition where the most efficient malware detection pipeline based upon robot learning wins!

Martin Vejmelka, one of the conferences principal speakers, will skill a machine learning system developed in Avast for hasty, repeated training of a hierarchical model upon a frightful database of files. This hierarchical model works by partitioning the character of executable files into clusters and automatically trains surgically remove local models in each cluster, leveraging both CPU and GPU training.

Machine Learning Prague’s goal is to foster discussion between machine learning practictioners and people who are interested int the applicatins of modern trends in artificial intelligence.