Model-based reinforcement learning: A survey

850 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Deep Reinforcement Learning (2023, unknown): “Reinforcement Learning (RL) discipline focuses on teaching agents how to make a series of decisions in an environment to maximize a reward signal [5], [6].”
Reinforcement Learning for Selective Key Applications in Power Systems: Recent Advances and Future Challenges (2021, unknown): ““Model-based” refers to the RL algorithms that explicitly estimate and online update an environment model from past observations and make decisions based on this model [16], such as upper conﬁdence RL (UCRL) [17] and Thompson sampling [18].”
Inventing Relational State and Action Abstractions for Effective and Efficient Bilevel Planning (2022, result): “Additional experimentation we conducted suggests that the GNN gets better with around an order of magnitude more data; this is consistent with previous ﬁndings that such strategies are very data-hungry [43].”
STMA: A Spatio-Temporal Memory Agent for Long-Horizon Embodied Task Planning (2025, unknown): “To address these challenges, researchers have proposed model-based RL [25, 26] to improve sample efficiency by constructing environment models, but these methods still struggle with model complexity and generalization.”