Multi-Objective Reinforcement Learning algorithms implementations.

Multi-objective Gymnasium environments for reinforcement learning

A Predator-Prey-Grass multi-agent gridworld environment implemented with Farama's Gymnasium, PettingZoo and MOMAland. Featuring dynamic spawning and deletion and partial observability of agents.

Multi-Objective Transport Network Design Problem (MO-TNDP) gymnasium environment.

Code for the paper Optimistic Linear Support and Successor Features as a Basis for Optimal Policy Transfer - ICML 2022

Multi-Objective Multi Agent Reinforcement Learning (MOMARL) experiments.