A challenge in multi-agent reinforcement learning is to be able to generalize over intractable state-action spaces. Inspired from Tesseract [Mahajan et al., 2021], this position paper investigates generalisation in state-action space over unexplored state-action pairs by modelling the transition and reward functions as tensors of low CP-rank. Initial experiments on synthetic MDPs show that using tensor decompositions in a model-based reinforcement learning algorithm can lead to much faster convergence if the true transition and reward functions are indeed of low rank.
@article{arxiv.2110.14524,
title = {Model based Multi-agent Reinforcement Learning with Tensor Decompositions},
author = {Pascal Van Der Vaart and Anuj Mahajan and Shimon Whiteson},
journal= {arXiv preprint arXiv:2110.14524},
year = {2021}
}