Log In Sign Up

Generating Adjacency-Constrained Subgoals in Hierarchical Reinforcement Learning

by   Tianren Zhang, et al.

Goal-conditioned hierarchical reinforcement learning (HRL) is a promising approach for scaling up reinforcement learning (RL) techniques. However, it often suffers from training inefficiency as the action space of the high-level, i.e., the goal space, is often large. Searching in a large goal space poses difficulties for both high-level subgoal generation and low-level policy learning. In this paper, we show that this problem can be effectively alleviated by restricting the high-level action space from the whole goal space to a k-step adjacency region centered by the current state using an adjacency constraint. We theoretically prove that the proposed adjacency constraint preserves the optimal hierarchical policy, and show that this constraint can be practically implemented by training an adjacency network that can discriminate between adjacent and non-adjacent subgoals. Experimental results on discrete and continuous control tasks show that our method outperforms the state-of-the-art HRL approaches.


page 7

page 8

page 13

page 18

page 19


Adjacency constraint for efficient hierarchical reinforcement learning

Goal-conditioned Hierarchical Reinforcement Learning (HRL) is a promisin...

Adversarially Guided Subgoal Generation for Hierarchical Reinforcement Learning

Hierarchical reinforcement learning (HRL) proposes to solve difficult ta...

Landmark-Guided Subgoal Generation in Hierarchical Reinforcement Learning

Goal-conditioned hierarchical reinforcement learning (HRL) has shown pro...

Adjacency-constrained hierarchical clustering of a band similarity matrix with application to Genomics

Motivation: Genomic data analyses such as Genome-Wide Association Studie...

Learning Goal Embeddings via Self-Play for Hierarchical Reinforcement Learning

In hierarchical reinforcement learning a major challenge is determining ...

Hierarchical Reinforcement Learning with Optimal Level Synchronization based on a Deep Generative Model

The high-dimensional or sparse reward task of a reinforcement learning (...

Spatially and Seamlessly Hierarchical Reinforcement Learning for State Space and Policy space in Autonomous Driving

Despite advances in hierarchical reinforcement learning, its application...

Code Repositories


PyTorch code accompanying the paper "Generating Adjacency-Constrained Subgoals in Hierarchical Reinforcement Learning" (NeurIPS 2020).

view repo