Machine Learning and Cyber Security Resources

[Faizann20]

http://fsecurify.com/machine-learning-and-cyber-security/

[kyrre]

More: http://www.covert.io/the-definitive-security-datascience-and...

[Faizann20 → kyrre]

Thanks.

[_5qp2]

Mandatory viewing for anybody considering messing with ML + infosec ... https://www.blackhat.com/html/webcast/02192015-secure-becaus...

[Faizann20]

Source: http://fsecurify.com/machine-learning-and-cyber-security/

[sctb → Faizann20]

Thanks, we've updated the link from http://www.kdnuggets.com/2017/01/machine-learning-cyber-secu....

[Faizann20 → sctb]

Thanks a lot.

[galapago]

VDiscover, a tool to predict vulnerable test cases using program traces, is missing: https://github.com/CIFASIS/VDiscover/

Full disclosure: I am one of the authors of "Toward large-scale vulnerability discovery using Machine Learning", where this tool was presented.

[Faizann20 → galapago]

Seems good. I'll add it.

[jcims]

I have a feeling that the inscrutability of deep learning models is going to provide a playground for fraud. There are some papers that I cannot find at the moment (Deep Flaws?) related to detection of model and input fouling via adversarial inputs.

I'm approaching ML from 20 years in infosec and am a bit undecided where I should focus my efforts. Using ML as a tool or defending ML against threats. Ultimately both will be applicable.

[vladsanchez]

I wondered about this topic last week! Whether it was possible to use ML/DL to model and predict cyber-attacks?

[Faizann20 → vladsanchez]

I hope you are not wondering now. :D

[vladsanchez → Faizann20]

Thanks for sharing, fascinating stuff in your link! :D