… This paper introduces a new scalable multi-objective deep reinforcementlearning (MODRL) framework based on deep Q-networks. We develop a high-performance MODRL framework …
… state of the art in MultiObjectiveReinforcementLearning (MORL) benchmarking for problems with … While this paper examined individual benchmarks, creating a larger benchmarksuite, …
… variant of Q-learning [26, 39, 54], learning a manifold of … also learning- or regularization-focused objectives such as staying close to a behavioral prior. Like MO-MPO, our multi-objective …
… Our key insight is to view constrained RL from a multi-objective perspective… multi-objective RL framework. We first formulate the constrained RL problem as a Constrained Multi-Objective …
Y Han, H Peng, C Mei, L Cao, C Deng, H Wang… - Knowledge-Based …, 2023 - Elsevier
… In the experimental section, benchmarksuites are used to test the performance of RLMMDE. Different types of advanced algorithms are used to verify the performance of the proposed …
… Through a number of proofs, the authors show the original DST problem to be quite basic, and not always representative of practical Multi-Objective Optimization (MOO) problems. In an …
… paper, a reinforcementlearning based adaptive operator selection method has been proposed for evolutionary multi-objective optimization. By using deep neural networks to learn the …
… , so we will first present a general discussion of multi-objective Q-learning before … As such, rather than evaluating a wide range of algorithms across the complete benchmarksuite, we …
… We propose a novel multi-objectivereinforcementlearning algorithm that successfully learns the optimal policy even for non-linear utility functions. Non-linear utility functions pose a …