A computationally lightweight safe learning algorithm
This addresses the problem of high computational cost in safe learning for physical systems, making it more applicable to embedded and high-dimensional settings, though it is incremental as it builds on existing safe learning frameworks.
The paper tackles the computational inefficiency of safe learning algorithms by proposing a new method that uses the Nadaraya-Watson estimator instead of Gaussian processes, achieving logarithmic scaling with data points and demonstrating results on a simulated robot manipulator.
Safety is an essential asset when learning control policies for physical systems, as violating safety constraints during training can lead to expensive hardware damage. In response to this need, the field of safe learning has emerged with algorithms that can provide probabilistic safety guarantees without knowledge of the underlying system dynamics. Those algorithms often rely on Gaussian process inference. Unfortunately, Gaussian process inference scales cubically with the number of data points, limiting applicability to high-dimensional and embedded systems. In this paper, we propose a safe learning algorithm that provides probabilistic safety guarantees but leverages the Nadaraya-Watson estimator instead of Gaussian processes. For the Nadaraya-Watson estimator, we can reach logarithmic scaling with the number of data points. We provide theoretical guarantees for the estimates, embed them into a safe learning algorithm, and show numerical experiments on a simulated seven-degrees-of-freedom robot manipulator.