Massively Scalable Inverse Reinforcement Learning in Google Maps

arXiv:2305.11290v4 Announce Type: replace
Abstract: Inverse reinforcement learning (IRL) offers a powerful and general framework for learning humans' latent preferences in route recommendation, yet no approach has successfully addressed planetary-scale problems with hundreds of millions of states and demonstration trajectories. In this paper, we introduce scaling techniques based on graph compression, spatial parallelization, and improved initialization conditions inspired by a connection to eigenvector algorithms. We revisit classic IRL methods in the routing context, and make the key observation that there …

abstract arxiv compression cs.lg framework general google graph humans maps paper recommendation reinforcement reinforcement learning route scalable scale scaling spatial type