Multi-objective Reinforcement Learning

Most real decisions involve trade-offs between goals such as speed and safety, cost and quality, or efficiency and fairness. Multi-objective reinforcement learning (MORL) studies how to train agents that learn to navigate these trade-offs and keep each goal as a separate objective. Our work develops algorithmic frameworks that combine evolutionary computation with reinforcement learning. These frameworks produce sets of policies that represent different trade-off solutions, with applications in logistics, finance, and autonomous systems. The project is funded by PAPIIT IA102025, UNAM.