Big Data & Machine Learning for network security: approaches and benchmarks
Room 2.31, 05/12/2018 (17:30-18:15)
The proposed session focuses on machine-learning and/or big-data approaches to network-based detection of and reaction to cybersecurity threats. With respect to this class of solutions, the session aims to present and discuss the issues to be faced and the different approaches available.
The objective is manifold.
First, to identify the most promising approaches.
Second, to solicit new ideas in this field.
Last but not least, to issue to the cybersecurity and machine-learning communities a call-for-action about the creation of a common public benchmarking approach (e.g. open data set of good/bad network traffic, labelled or not) because currently comparing results from different algorithms/tools is nearly impossible, as each author is evaluating performance over its own custom dataset.
Organised by: Antonio LIOY (Politecnico di Torino, DAUIN, Italy)
Comments are closed