Reducing hardware experiments for model learning and policy optimization

25 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.

Sim-to-Real: Learning Agile Locomotion For Quadruped Robots (2018, methodology): “[14] used Gaussian Processes, a non-parametric model, to minimize the error between simulation and real physics.”
Preparing for the Unknown: Learning a Universal Policy with Online System Identification (2017, methodology): “Some widely-used models, such as linear models [3], Gaussian processes [8, 11], and differential equations [35, 4], have proven effective for continuous dynamics and control tasks with relatively low action space.”
PAC Reinforcement Learning With an Imperfect Model (2018, result): “In another related work, (Ha and Yamane 2015) propose an algorithm for linear control problems that is similar in spirit to our Algorithm 1.”
Planning and Execution using Inaccurate Models with Provable Guarantees (2020, background): “been shown to work well in episodic settings [26, 12, 27].”
Simulation-based design of dynamic controllers for humanoid balancing (2016, methodology): “Widely used models include state transition functions [5], linear models [5], Gaussian processes [6], [7], and differential equations [8], [9].”