19 June - AIAI/IPAB Seminar - Stefano Albrecht

Link

https://eu.bbcollab.com/guest/97e5b2666aa74b30870a4c83ad03760a  

Speaker

Stefano Albrecht

Title

Our submissions to NeurIPS 2020 on reinforcement learning in multi-agent systems

Abstract

Our group specialises in algorithm development for autonomous decision making in complex environments, with a focus on deep reinforcement learning applied to multi-agent interaction problems. I will give brief summaries of our 5 new submissions to NeurIPS 2020 which, among others, address following problems: efficient multi-agent learning in environments with sparse rewards; modelling other agents using only local/partial observations; learning to collaborate in open teams in which other agents may enter or leave at any time; and quantum-robust authentication schemes based on abstract multi-agent interaction.