Learning Trust Over Directed Graphs in Multiagent Systems

Orhan Eren Akgun, Arif Kerem Dayi, Stephanie Gil, Angelia Nedich
Proceedings of The 5th Annual Learning for Dynamics and Control Conference, PMLR 211:142-154, 2023.

Abstract

We address the problem of learning the legitimacy of other agents in a multiagent network when an unknown subset is comprised of malicious actors. We specifically derive results for the case of directed graphs and where stochastic side information, or observations of trust, is available. We refer to this as “learning trust” since agents must identify which neighbors in the network are reliable, and we derive a protocol to achieve this. We also provide analytical results showing that under this protocol i) agents can learn the legitimacy of all other agents almost surely, and that ii) the opinions of the agents converge in mean to the true legitimacy of all other agents in the network. Lastly, we provide numerical studies showing that our convergence results hold in practice for various network topologies and variations in the number of malicious agents in the network.

Cite this Paper

BibTeX
@InProceedings{pmlr-v211-akgun23a, title = {Learning Trust Over Directed Graphs in Multiagent Systems}, author = {Akgun, Orhan Eren and Dayi, Arif Kerem and Gil, Stephanie and Nedich, Angelia}, booktitle = {Proceedings of The 5th Annual Learning for Dynamics and Control Conference}, pages = {142--154}, year = {2023}, editor = {Matni, Nikolai and Morari, Manfred and Pappas, George J.}, volume = {211}, series = {Proceedings of Machine Learning Research}, month = {15--16 Jun}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v211/akgun23a/akgun23a.pdf}, url = {https://proceedings.mlr.press/v211/akgun23a.html}, abstract = {We address the problem of learning the legitimacy of other agents in a multiagent network when an unknown subset is comprised of malicious actors. We specifically derive results for the case of directed graphs and where stochastic side information, or observations of trust, is available. We refer to this as “learning trust” since agents must identify which neighbors in the network are reliable, and we derive a protocol to achieve this. We also provide analytical results showing that under this protocol i) agents can learn the legitimacy of all other agents almost surely, and that ii) the opinions of the agents converge in mean to the true legitimacy of all other agents in the network. Lastly, we provide numerical studies showing that our convergence results hold in practice for various network topologies and variations in the number of malicious agents in the network.} }
Endnote
%0 Conference Paper %T Learning Trust Over Directed Graphs in Multiagent Systems %A Orhan Eren Akgun %A Arif Kerem Dayi %A Stephanie Gil %A Angelia Nedich %B Proceedings of The 5th Annual Learning for Dynamics and Control Conference %C Proceedings of Machine Learning Research %D 2023 %E Nikolai Matni %E Manfred Morari %E George J. Pappas %F pmlr-v211-akgun23a %I PMLR %P 142--154 %U https://proceedings.mlr.press/v211/akgun23a.html %V 211 %X We address the problem of learning the legitimacy of other agents in a multiagent network when an unknown subset is comprised of malicious actors. We specifically derive results for the case of directed graphs and where stochastic side information, or observations of trust, is available. We refer to this as “learning trust” since agents must identify which neighbors in the network are reliable, and we derive a protocol to achieve this. We also provide analytical results showing that under this protocol i) agents can learn the legitimacy of all other agents almost surely, and that ii) the opinions of the agents converge in mean to the true legitimacy of all other agents in the network. Lastly, we provide numerical studies showing that our convergence results hold in practice for various network topologies and variations in the number of malicious agents in the network.
APA
Akgun, O.E., Dayi, A.K., Gil, S. & Nedich, A.. (2023). Learning Trust Over Directed Graphs in Multiagent Systems. Proceedings of The 5th Annual Learning for Dynamics and Control Conference, in Proceedings of Machine Learning Research 211:142-154 Available from https://proceedings.mlr.press/v211/akgun23a.html.

Related Material