Hierarchical Universal Value Function Approximators

Rushiv Arora

There have been key advancements to building universal approximators for multi-goal collections of reinforcement learning value functions -- key elements in estimating long-term re…