Hi Egon, Yes that's going in the same direction. I just wanted to share this interesting paper which can help us for implementing. I am currently studying Machine Learning for this purpose and hope to be able to contribute soon :)
Best regards On Fri, 5 Aug 2016 at 9:07 PM, Egon Kidmose <[email protected]> wrote: > Hey Franck, > > Nice reads, thanks. > > The topic has been up before: > > https://mail-archives.apache.org/mod_mbox/incubator-metron-dev/201606.mbox/%3C8FE3632E-1B91-4C66-9CE4-578D609768B6%40cisco.com%3E > (and as I skimmed it a saw you also were in on that thread...) > > I had exactly this scenario in mind when I added some of the stories on > Yazan Boshmaf's document here: https://goo.gl/QAxiH6 > Does this cover what you are envisioning? > > > > > Mvh. / BR > Egon Kidmose > > On Fri, Aug 5, 2016 at 2:27 AM, Franck Vervial <[email protected]> wrote: > > > Hello, > > > > An interesting article from MIT: > > http://news.mit.edu/2016/ai-system-predicts-85-percent- > > cyber-attacks-using-input-human-experts-0418 > > > > AI2 paper: https://people.csail.mit.edu/kalyan/AI2_Paper.pdf > > > > Combining unsupervised machine learning and supervised machine learning > > with the help of human SOC analysts will help to reduce consequently the > > number of False Positives. > > > > As Apache Metron is the future of SIEM, I think this is an avenue to > > explore. > > > > Franck > > >
