Future Aware Safe Active Learning of Time Varying Systems using Gaussian Processes
Not on the map. Why?
Abstract (arXiv)
Experimental exploration of high-cost systems with safety constraints, common in engineering applications, is a challenging endeavor. Data-driven models offer a promising solution, but acquiring the requisite data remains expensive and is potentially unsafe. Safe active learning techniques prove essential, enabling the learning of high-quality models with minimal expensive data points and high safety. This paper introduces a safe active learning framework tailored for time-varying systems, addressing drift, seasonal changes, and complexities due to dynamic behavior. The proposed Time-aware Integrated Mean Squared Prediction Error (T-IMSPE) method minimizes posterior variance over current and future states, optimizing information gathering also in the time domain. Empirical results highlight T-IMSPE's advantages in model quality through toy and real-world examples. State of the art Gaussian processes are compatible with T-IMSPE. Our theoretical contributions include a clear delineation which Gaussian process kernels, domains, and weighting measures are suitable for T-IMSPE and even beyond for its non-time aware predecessor IMSPE.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 39 of its 97 references that are in the map. First posted on arXiv 17 May 2024.
Builds on
39 of 97 references.
References · 39
- 2018GPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU Accelerationcited by 609
- 2019BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learningcited by 432
- 2017A General Safety Framework for Learning-Based Control in Uncertain Robotic Systemscited by 163
- 2016Bayesian Optimization with Safety Constraints: Safe and Automatic Parameter Tuning in Roboticscited by 134
- 2019Online Control with Adversarial Disturbancescited by 92
- 2016Variational Fourier features for Gaussian processescited by 85
- 2018Stagewise Safe Bayesian Optimization with Gaussian Processescited by 79
- 2015Safe Controller Optimization for Quadrotors with Gaussian Processescited by 78
- 2020A Survey of Reinforcement Learning Algorithms for Dynamically Varying Environmentscited by 62
- 2019Safe Exploration for Interactive Machine Learningcited by 51
- 2014Variational Gaussian Process State-Space Modelscited by 43
- 2020Active Learning for Deep Gaussian Process Surrogatescited by 38
27 more
- 2021Non-stationary Online Learning with Memory and Non-stochastic Controlcited by 33
- 2017Identification of Gaussian Process State Space Modelscited by 31
- 2017Replication or exploration? Sequential design for stochastic simulation experimentscited by 30
- 2020SAMBA: Safe Model-Based & Active Reinforcement Learningcited by 30
- 2021GoSafe: Globally Optimal Safe Robot Learningcited by 27
- 2014Exploration vs Exploitation vs Safety: Risk-averse Multi-Armed Banditscited by 26
- 2019Actively Learning Gaussian Process Dynamicscited by 26
- 2022Bayesian Active Learning with Fully Bayesian Gaussian Processescited by 26
- 2019Safe Contextual Bayesian Optimization for Sustainable Room Temperature PID Control Tuningcited by 24
- 2022Batch Multi-Fidelity Active Learning with Budget Constraintscited by 20
- 2022Gaussian Process Priors for Systems of Linear Partial Differential Equations with Constant Coefficientscited by 17
- 2024On Safety in Safe Bayesian Optimizationcited by 15
- 2022Constraining Gaussian Processes to Systems of Linear Ordinary Differential Equationscited by 14
- 2022Safe Active Learning for Multi-Output Gaussian Processescited by 13
- 2022On Controller Tuning with Time-Varying Bayesian Optimizationcited by 13
- 2024Safe Active Learning for Time-Series Modeling with Gaussian Processescited by 12
- 2021Partitioned Active Learning for Heterogeneous Systemscited by 10
- 2023Amortized Inference for Gaussian Process Hyperparameters of Structured Kernelscited by 10
- 2022Structural Kernel Search via Bayesian Optimization and Symbolical Optimal Transportcited by 8
- 2019A Learnable Safety Measurecited by 7
- 2022NCVX: A General-Purpose Optimization Solver for Constrained Machine and Deep Learningcited by 7
- 2021Failure-averse Active Learning for Physics-constrained Systemscited by 5
- 2022Designing Robust Biotechnological Processes Regarding Variabilities using Multi-Objective Optimization Applied to a Biopharmaceutical Seed Train Designcited by 4
- 2018Known Boundary Emulation of Complex Computer Modelscited by 2
- 2020Data selection for multi-task learning under dynamic constraintscited by 2
- 2021Active Learning in Gaussian Process State Space Modelcited by 2
- 2024Efficiently Computable Safety Bounds for Gaussian Processes in Active Learningcited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.