TreeQN and ATreeC: Differentiable Tree-Structured Models for Deep Reinforcement Learning

Abstract

Combining deep model-free reinforcement learning with on-line planning is a promising approach to building on the successes of deep RL. On-line planning with look-ahead trees has proven successful in environments where transition models are known a priori. However, in complex environments where transition models need to be learned from data, the deficiencies of learned models have limited their utility for planning. To address these challenges, we propose TreeQN, a differentiable, recursive, tree-structured model that serves as a drop-in replacement for any value function network in deep RL with discrete actions. TreeQN dynamically constructs a tree by recursively applying a transition model in a learned abstract state space and then aggregating predicted rewards and state-values using a tree backup to estimate Q-values. We also propose ATreeC, an actor-critic variant that augments TreeQN with a softmax layer to form a stochastic policy network. Both approaches are trained end-to-end, such that the learned model is optimised for its actual use in the tree. We show that TreeQN and ATreeC outperform n-step DQN and A2C on a box-pushing task, as well as n-step DQN and value prediction networks (Oh et al. 2017) on multiple Atari games. Furthermore, we present ablation studies that demonstrate the effect of different auxiliary losses on learning transition models.

Cited by in corpus (17)

• Universal Planning Networks
• MPC-Inspired Neural Network Policies for Sequential Decision Making
• Risks from Learned Optimization in Advanced Machine Learning Systems
• Muesli: Combining Improvements in Policy Optimization
• Fast Efficient Hyperparameter Tuning for Policy Gradients
• The Value Equivalence Principle for Model-Based Reinforcement Learning
• Deep Residual Reinforcement Learning
• Plan2Vec: Unsupervised Representation Learning by Latent Plans
• Toybox: A Suite of Environments for Experimental Evaluation of Deep Reinforcement Learning
• XLVIN: eXecuted Latent Value Iteration Nets
• Integrating Algorithmic Planning and Deep Learning for Partially Observable Navigation
• Influence-Augmented Online Planning for Complex Environments
• Visualizing MuZero Models
• Self-Consistent Models and Values
• Differentiable Robust LQR Layers
• Proper Value Equivalence
• Neural Algorithmic Reasoners are Implicit Planners