Multi-Objective Hierarchical Reinforcement Learning

Value Function Decomposition

Technique that decomposes the global value function into contributions from each subtask and objective, facilitating distributed learning in hierarchies.

← Zurück