A Multi-Objective Deep Reinforcement Learning Framework

Abstract

This paper introduces a new scalable multi-objective deep reinforcement learning (MODRL) framework based on deep Q-networks. We develop a high-performance MODRL framework that supports both single-policy and multi-policy strategies, as well as both linear and non-linear approaches to action selection. The experimental results on two benchmark problems (two-objective deep sea treasure environment and three-objective Mountain Car problem) indicate that the proposed framework is able to find the Pareto-optimal solutions effectively. The proposed framework is generic and highly modularized, which allows the integration of different deep reinforcement learning algorithms in different complex problem domains. This therefore overcomes many disadvantages involved with standard multi-objective reinforcement learning methods in the current literature. The proposed framework acts as a testbed platform that accelerates the development of MODRL for solving increasingly complicated multi-objective problems.

21 pages

References in corpus (9)

• TensorFlow: A system for large-scale machine learning
• Benchmarking Deep Reinforcement Learning for Continuous Control
• A Survey of Multi-Objective Sequential Decision-Making
• Multi-Objective Deep Reinforcement Learning
• Particle Swarm Optimization for Generating Interpretable Fuzzy Reinforcement Learning Policies
• Dynamic Weights in Multi-Objective Deep Reinforcement Learning
• Online Reinforcement Learning in Stochastic Games
• Two geometric input transformation methods for fast online reinforcement learning with neural nets
• Online Reinforcement Learning for Real-Time Exploration in Continuous State and Action Markov Decision Processes

Cited by in corpus (15)

• Deep Reinforcement Learning for Multi-Agent Systems: A Review of Challenges, Solutions and Applications
• Deep Reinforcement Learning for Cyber Security
• Exploration in Deep Reinforcement Learning: From Single-Agent to Multiagent Domain
• Multi-Objective Multi-Agent Decision Making: A Utility-based Analysis and Survey
• Levels of explainable artificial intelligence for human-aligned conversational explanations
• Deep Reinforcement Learning for Radio Resource Allocation and Management in Next Generation Heterogeneous Wireless Networks: A Survey
• A Practical Guide to Multi-Objective Reinforcement Learning and Planning
• Solving the scalarization issues of Advantage-based Reinforcement Learning Algorithms
• Deep W-Networks: Solving Multi-Objective Optimisation Problems With Deep Reinforcement Learning
• Online Multi-Objective Model-Independent Adaptive Tracking Mechanism for Dynamical Systems
• Reinforcement Learning for Joint Optimization of Multiple Rewards
• Relationship Explainable Multi-objective Reinforcement Learning with Semantic Explainability Generation
• Relationship Explainable Multi-objective Optimization Via Vector Value Function Based Reinforcement Learning
• Review, Analysis and Design of a Comprehensive Deep Reinforcement Learning Framework
• Prioritized Soft Q-Decomposition for Lexicographic Reinforcement Learning