Safe Exploration for Optimization with Gaussian Processes

By Yanan Sui, Alkis Gotovos, Joel W. Burdick, Andreas Krause. 2015

461 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.

Citations in the corpus

Safe Model-based Reinforcement Learning with Stability Guarantees (2017, background): “The algorithms in [17] and [18] provide high-probability safety guarantees for any parameter that is evaluated on the real system.”
Trial without Error: Towards Safe Reinforcement Learning via Human Intervention (2017, background): “Related approaches have been pursued in recent work on safe exploration [26].”
A Review of Safe Reinforcement Learning: Methods, Theory and Applications (2022, methodology): “In a Gaussian Process (GP) of model-free settings, Sui et al [260] use a GP method to present the unknown reward function from noise samples, the exploration by leveraging a GP method is improved to reduce uncertainty and ensure agent safety, more particularly, the GP method is used to predict unkno”
Constrained Efficient Global Optimization of Expensive Black-box Functions (2022, methodology): “This is due to the possibility of getting stuck in a local minimum for SafeOPT.”
A tutorial on Gaussian process regression: Modelling, exploring, and exploiting functions (2016, background): “For further technical details, we refer the interested reader to Sui et al. (2015).”
Safe Exploration in Finite Markov Decision Processes with Gaussian Processes (2016, background): “We known that l1(s)− Ld(s, s) = l1(s) ≥ l0(s) ≥ h, where the last inequality follows from Lemma 3.”
Safe Learning in Robotics: From Learning-Based Control to Safe Reinforcement Learning (2021, background): “1 Nonlinear (69, 79, 82), GP (78, 79, 80, 81) NN (70, 75, 85, 86, 87), Tabular (73) Input & state (69, 70, 73, 75, 78, 79, 80, 81, 82) Trajectory (87) Input & state (69, 70, 73, 75, 78, 79, 80, 81, 82) Trajectory (85, 86, 87) Stabilization (81), Tracking (79), Navigation (69, 70, 73, 82, 85, 86), Lo”
A Review of Safe Reinforcement Learning: Methods, Theories, and Applications (2024, unknown): “In a GP of model-free settings, Sui et al. [44] use a GP method to present the unknown reward function from noise samples, the exploration by leveraging a GP method is improved to reduce uncertainty and ensure agent safety.”
Conservative Bandits (2016, background): “Another line of work considers safe exploration in the related context of optimization (Sui et al., 2015).”
On Kernelized Multi-armed Bandits (2017, background): “), GP-EST Wang et al. (2016) and GP-UCB, including the contextual Krause and Ong (2011), high-dimensional Djolonga et al. (2013); Wang et al. (2013), time-varying Bogunovic et al. (2016) safety-aware Gotovos et al. (2015), budget-constraint Hoffman et al. (2013) and noise-free De Freitas et al. (201”
Safe controller optimization for quadrotors with Gaussian processes (2015, methodology): “In the next section, we modify the algorithm in [14] to use the GP’s prediction directly in order to estimate these sets.”
Bayesian optimization with safety constraints: safe and automatic parameter tuning in robotics (2016, unknown): “However, the algorithm of Sui et al. (2015) considers safety as a minimum performance requirement.”