Palo Alto Networks offers access to cyber security data sets.
The data sets are described in the paper "Machine Learning in Cyber-Security - Problems, Challenges and Data Sets". We present important cyber security problems.
We explain why text book machine learning fails to cope with them.
The machine learning challenges are of interest to the machine learning community, not only to the cyber security community.
We also provide the data sets for academic use. One of the data sets was crated merging data from Palo Alto Networks and Shodan (shodan.io). For details and access to the data, see https://arxiv.org/abs/1812.07858 |