… action selection purposes in on-line reinforcementlearning. Instead, we propose a novel non-… previous work on multi-objective optimization and reinforcement learning. Furthermore, we …
K Van Moffaert, A Nowé - The Journal of Machine Learning Research, 2014 - jmlr.org
… multi-objective optimization problems. In our work, we focus … reinforcementlearning for multi-objective problems. Reinforcement learning (Sutton and Barto, 1998) is a machine learning …
… algorithm for multi-objectivereinforcementlearning (MORL) with linear preferences, with the goal of enabling few-shot adaptation to new tasks. In MORL, the aim is to learn policies over …
… We argue that multi-objective methods are underrepresented … justify the need for explicitly multi-objective approaches. Key to … of multi-objective RL, ie, that the appropriate multi-objective …
… a new scalable multi-objective deep reinforcementlearning (… of different deep reinforcement learning algorithms in different … with standard multi-objectivereinforcementlearning methods …
F Zou, GG Yen, L Tang, C Wang - Information Sciences, 2021 - Elsevier
… In this paper, a reinforcementlearning-based dynamic multi-objective evolutionary algorithm, called RL-DMOEA, which seamlessly integrates reinforcementlearning framework and …
… As a baseline, we use a basic Multi-Objective DQN approach (MO); a single multi-objective DQN continuously trained on only the current w through scalarized Deep Q-learning. MO …
… using reinforcementlearning strategies based on a novel prediction-guided evolutionary learning … In order to benchmark our proposed algorithm, we design a set of multi-objective robot …