Machine Learning for Cybersecurity with Evan Wright

800 800 The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)

This week my guest is Evan Wright, principal data scientist at cybersecurity startup Anomali. In my interview with Evan, he and I discussed about a number of topics surrounding the use of machine learning in cybersecurity.

Future of Data Summit

If Evan’s name sounds familiar, it’s because Evan was the winner of the O’Reilly Strata+Hadoop World ticket giveaway earlier this month. We met up at the conference last week and took advantage of the opportunity to record this show.

Our conversation covers, among other topics, the three big problems in cybersecurity that ML can help out with, the challenges of acquiring ground truth in cybersecurity and some ways to accomplish it, and the use of decision trees, generative adversarial networks, and other algorithms in the field.

Get Your Stickers

This Week in Machine Learning & AI StickerYep, we’ve got some nice new stickers for the podcast, and we want to give you one! We’re continuing the sticker contest we kicked off last week. Send us your favorite quote from today’s show via a comment or post on Facebook, Twitter, Youtube or SoundCloud, as well as via the show notes page, and we’ll send a sticker your way!

About Evan Wright

Mentioned in the Interview

  • Shawn Wang

    Model interpretability: “The biggest innovations in the cybersecurity space are being able to interpret a model to go back to the data collection, make suggestions of new data to collect, and then improve the process and start all over again.”

    This is a massive startup idea right here and it is a generalizable problem beyond just this domain. I hope someone is working on this and if someone is I hope you interview him/her!

  • Garrett

    “There’s this question of if you tell an adversary you can detect them, what will happen?” Seems like a pretty difficult dilemma in the cyber security domain. I think I side a bit more with Evan on this one; there’s got to be a better way to share defensive strategies that doesn’t involve easy access by attackers.

  • Rob

    Evan talks about the multiple parameters that must be tuned when using something like XGBoost. I hear everyone in the Deep Learning community talking about grid searches, etc, for parameter tuning. What’s keeping everyone back from the use of Evolutionary/Swarm algorithms that have shown great promise in combinatorial search?

  • Nick

    Great show! “It would be helpful if people in the field would spend some time on types of overfitting.” This is an interesting point. I haven’t seen much research on how different methods of regularization and augmentation impact types of overfitting. This is really important work.

Leave a Reply

Your email address will not be published.