Abstract
This work addresses a specific case of Goal Recognition (GR), where
a malicious actor (the attacker) seeks to reach and damage one of
several sensitive targets, while the observer (the defender) must
identify the attacker’s target and allocate limited resources to pro-
tect it. Focusing on real-world physical and cyber security scenarios,
the defender faces a trade-off between acting early, with limited in-
formation, or waiting for more data but risking insufficient time to
defend. Our contributions include introducing a game-theoretic for-
mulation of this instance of GR, which captures the time-sensitive
nature of these scenarios, and providing an efficient method to
compute Nash equilibria using the fictitious play learning scheme.
Experimental results confirm that our method equips the defend
a malicious actor (the attacker) seeks to reach and damage one of
several sensitive targets, while the observer (the defender) must
identify the attacker’s target and allocate limited resources to pro-
tect it. Focusing on real-world physical and cyber security scenarios,
the defender faces a trade-off between acting early, with limited in-
formation, or waiting for more data but risking insufficient time to
defend. Our contributions include introducing a game-theoretic for-
mulation of this instance of GR, which captures the time-sensitive
nature of these scenarios, and providing an efficient method to
compute Nash equilibria using the fictitious play learning scheme.
Experimental results confirm that our method equips the defend
| Original language | English |
|---|---|
| Title of host publication | 24th International Conference on Autonomous Agents and Multiagent Systems |
| Publication status | Accepted/In press - 19 Dec 2024 |
Keywords
- obfuscation legibility pathfinding