Have you ever told a secret to a friend and later regretted it?

At #NDSS23, Alex will present our approach to machine unlearning of features and labels. At least for learning models, we can remove information in retrospection🫠.

Unlike previous work, which is limited to forgetting points, we can unlearn arbitrary feature values from a trained model. This is possible because we estimate their influence and compute a reverse model update (like a hangover🍺)

Check out our paper https://www.mlsec.org/docs/2023-ndss.pdf

Are you still writing your space and satellite security papers for #SpaceSec23? The deadline has been extended to 13 January AoE, so almost one week to get that paper in!

CfP: https://easychair.org/cfp/SpaceSec23
#NDSS23: https://www.ndss-symposium.org/ndss2023/submissions/cfp-spacesec/

Only 1 day left to submit to the #NDSS23 MadWeb workshop: https://madweb.work

https://twitter.com/NDSSSymposium/status/1611048327289704471

📣​ Only a few weeks left to submit your awesome binary analysis research to the Binary Analysis Research Workshop (#BAR23, co-located with #NDSS23)! Looking forward to your submissions!

https://bar2023.github.io
(deadline: January 3rd AoE)

Tactics, Threats, & Targets: Our
#NDSS23 paper presents a #CyberSecurity inspired framework to characterize the threat of coordinated #disinformation operations, with insights informed by the experiences of mitigation experts.

https://www.shujaatmirza.com/papers/disinfo_threat_ndss23.pdf