no code implementations • 27 Nov 2023 • John Kliem, Prithviraj Dasgupta
We investigate the effect of reward shaping in improving the performance of reinforcement learning in the context of the real-time strategy, capture-the-flag game.
no code implementations • 7 Jun 2023 • Prithviraj Dasgupta
We propose a novel technique that can identify parts of demonstrated trajectories that have not been significantly modified by the adversary and utilize them for learning, using temporally extended policies or options.
no code implementations • 14 Apr 2023 • Bryan Brandt, Prithviraj Dasgupta
We propose a novel algorithm that can generate synthetic, human-like, decision making data while starting from a very small set of decision making data collected from humans.
no code implementations • 22 Nov 2021 • Prithviraj Dasgupta, Zachariah Osman
We consider the problem of generating adversarial malware by a cyber-attacker where the attacker's task is to strategically modify certain bytes within existing binary malware files, so that the modified files are able to evade a malware detector such as machine learning-based malware classifier.
no code implementations • 10 Feb 2020 • Prithviraj Dasgupta, Joseph B. Collins, Michael McCarrick
The objective of the adversary is to evade the learner's prediction mechanism by sending adversarial queries that result in erroneous class prediction by the learner, while the learner's objective is to reduce the incorrect prediction of these adversarial queries without degrading the prediction quality of clean queries.
no code implementations • 4 Dec 2019 • Prithviraj Dasgupta, Joseph B. Collins
A critical vulnerability of these algorithms is that they are susceptible to adversarial attacks where a malicious entity called an adversary deliberately alters the training data to misguide the learning algorithm into making classification errors.