In multi-agent systems, agents often have to rely on interactions with other agents in order to accomplish a given task. They hence need to assess the trustworthiness of other agents, which is particularly difficult if the latter change their behavior dynamically. The two common techniques to solve this problem are Hidden Markov Models (HMMs) and standard Beta Reputation Systems (BRS) equipped with a simple decay mechanism to discount older interactions. We propose instead to use \emph{Page-Hinkley} statistics in BRS to detect and dismiss an agent whose behavior worsens. Our experimental study demonstrates that our method outperforms HMMs and, in the vast majority of tested settings, either outperforms or is on par with other typically used BRS-type methods.
Article ID: 2022L4
Month: May
Year: 2022
Address: Online
Venue: Canadian Conference on Artificial Intelligence
Publisher: Canadian Artificial Intelligence Association
URL: https://caiac.pubpub.org/pub/op2o477r