Our conference paper “Intrusion Tolerance for Networked Systems Through Two-Level Feedback Control” was accepted to DSN24. A preprint will be available soon.
A preprint of our paper “Automated Security Response through Online Learning with Adaptive Conjectures” is available.
29 Dec ‘23
The camera ready of our paper “Scalable Learning of Intrusion Responses through Recursive Decomposition” is now available in the GameSec23 proceedings.
23 Dec ‘23
Our conference paper “Online Policy Adaptation for Networked Systems using Rollout” was accepted to NOMS24. A preprint will be available soon.
19 Dec ‘22
Dr. Neil Dhir from Siemens Research is visiting our lab at KTH.
Our system the “Cyber Security Learning Environment” (CSLE) is now available open source here.
26 Aug ‘23
Our conference paper “Scalable Learning of Intrusion Responses through Recursive Decomposition” was accepted to GameSec23. A preprint will be available soon.
24 June ‘23
Our conference paper “Digital Twins for Security Automation” is available in the NOMS´23 proceedings.
24 June ‘23
Our conference paper “Demonstrating a System for Dynamically Meeting Management Objectives on a Service Mesh” is available in the NOMS´23 proceedings.
Dr. Neil Dhir from Siemens Research is visiting us at KTH.
6 Dec ‘22
Dr. Erik Hemberg from MIT ALFA is visiting us at KTH.
1 Nov ‘22
Attending CNSM´22 in Thessaloniki, Greece.
27 Aug ‘22
Our conference paper “An Online Framework for Adapting Security Policies in Dynamic IT Environments” was accepted to CNSM´22.
17 Jul ‘22
Attending ICML 22 in Baltimore. We will present our paper “Learning Security Strategies through Game Play and Optimal Stopping (Camera ready version) on 22 July at the workshop on machine learning for cyber security (workshop). A pre-recorded video of our presentation is available.
A preprint of our paper “Learning Security Strategies through Game Play and Optimal Stopping” is available. To be submitted.
24 May ‘22
We presented a poster at the CDIS spring conference at KTH.
22 May ‘22
Our journal paper “Intrusion Prevention through Optimal Stopping” was accepted to IEEE Transactions on Network and Service Management and is available in early access.
30 Apr ‘22
Our demonstration at IEEE/IFIP Network Operations and Management Symposium was awarded the best demonstration award. A preprint is available here and a video of the demonstration is available here.
14 Apr ‘22
A demonstration of our method for finding effective security strategies through reinforcement learning is available.
11 Mar ‘22
Our demo paper “A System for Interactive Examination of Learned Security Policies” was accepted to NOMS´22, a preprint is available.
17 Jan ‘22
We had a poster presentation at the Digital Futures Machine Learning day at KTH.
30 Oct ‘21
Preprint of our journal paper “Intrusion Prevention through Optimal Stopping” is available.
26 Oct ‘21
Our conference paper “Learning Intrusion Prevention Policies through Optimal Stopping” is available in the CNSM´21 proceedings.
21 Oct ‘21
Master thesis Bayesian Reinforcement Learning Methods for Network Intrusion Prevention by Antonio Frederico Nesti Lopes, available here.
15 Oct ‘21
Attended the CDIS Fall Research Workshop at Balingsholm and presented our research project. The slides are available here.
30 Aug ‘21
Antonio Frederico Nesti Lopes, master thesis student in our group, successfully defended his M.Sc., with thesis entitled “Bayesian Reinforcement Learning Methods for Network Intrusion Prevention”. The camera ready version will be available soon.
23 July ‘21
Attending the DLRL summer school and presenting a poster. Our poster is available here.
16 July ‘21
Our conference paper “Learning Intrusion Prevention Policies through Optimal Stopping” was accepted to CNSM´21, a preprint is available.
15 June ‘21
Preprint of our conference paper “Learning Intrusion Prevention Policies through Optimal Stopping” is available.
24 Mar ‘21
Slides of our presentation at the CDIS spring conference is available.
14 Sep ‘20
Our CNSM´20 paper “Finding Effective Security Strategies through Reinforcement Learning and Self-Play” was accepted, a preprint is available.
