Malware classification meets crowd sourcing (sponsor presentation)

Presented at VB2015, Oct. 1, 2015, 4:30 p.m. (30 minutes).

Early this year, *Microsoft Malware Protection Center* and *Microsoft Azure Machine Learning Group* released a large corpus of malware data to a public competition. Since this was an interesting competition, the results of which could benefit the whole anti-malware industry, *HP Security Research* decided to participate. We would like to present the difficulties of the challenge, and to describe the machine-learning techniques that work well and those that do not work well in malware classification.

Presenters:

• John Park - HP
  John Park John is a senior researcher at HP Security Research - Applied Data Science team. His research interest lies in explaining computer security in layman's term and in wholistic view, and applying machine learning to detect and react to the threats faster. In his downtime, he likes to relax while machines are crunching some numbers, or mentor people on data science, or compete in hackathons. He has a dual-degree of EECS and Cognitive Science, from UC Berkeley.

Links:

• https://www.virusbulletin.com/conference/vb2015/abstracts/malware-classification-meets-crowd-sourcing

Similar Presentations:

• ToorCon: Seattle 2008 / Malware Classification using Machine Learning
• DerbyCon 3.0 All in the Family (2013) / The Malware Management Framework, a process you can use to find advanced malware. We found WinNTI with it!
• DEF CON 32 (2024) / Master Class: Hands-On Machine Learning to Enhance Malware Analysis, Classification, and Detection Workshop