arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.09033cs.LGstat.ML
收藏

有界支撑噪声下高斯过程回归的误差界及其在安全认证中的应用

Error Bounds For Gaussian Process Regression Under Bounded Support Noise With Applications To Safety Certification

University of Colorado Boulder(科罗拉多大学博尔德分校);Delft University of Technology(代尔夫特理工大学)

Robert Reed, Luca Laurenti, Morteza Lahijanian

AI总结:

本文针对有界支撑噪声下的高斯过程回归(GPR),利用集中不等式和RKHS复杂度假设推导出概率与确定性误差界,并证明其在深度核学习(DKL)中表现优异,结合随机障碍函数显著提升了未知动态系统的安全概率。

摘要:

高斯过程回归(GPR)是一种强大且优雅的方法,用于从噪声数据中学习复杂函数,具有广泛的应用范围,包括在安全关键领域。此类应用具有两个关键特征:(i)它们需要严格的误差量化,以及(ii)由于物理约束等原因,噪声通常是有界的且非高斯的。虽然存在在非高斯噪声下应用GPR的误差界,但它们在实践中往往过于严格和保守。在本文中,我们为有界支撑噪声下的GPR提供了新的误差界。具体而言,通过依赖集中不等式并假设潜函数在与GP核对应的再生核希尔伯特空间(RKHS)中具有低复杂度,我们推导出了GPR误差的概率界和确定性界。我们表明,这些误差界比现有的最先进误差界要严格得多,并且特别适合于具有神经网络核的GPR,即深度核学习(DKL)。此外,受安全关键领域应用的启发,我们说明了如何将这些误差界与随机障碍函数相结合,以成功地从有限数据中量化未知动态系统的安全概率。我们通过几个基准测试和与现有误差界的比较验证了我们方法的有效性。结果表明,我们的误差界始终更小,并且DKL可以产生比样本噪声更严格的误差界,从而显著提高控制系统的安全概率。

英文摘要:

Gaussian Process Regression (GPR) is a powerful and elegant method for learning complex functions from noisy data with a wide range of applications, including in safety-critical domains. Such applications have two key features: (i) they require rigorous error quantification, and (ii) the noise is often bounded and non-Gaussian due to, e.g., physical constraints. While error bounds for applying GPR in the presence of non-Gaussian noise exist, they tend to be overly restrictive and conservative in practice. In this paper, we provide novel error bounds for GPR under bounded support noise. Specifically, by relying on concentration inequalities and assuming that the latent function has low complexity in the reproducing kernel Hilbert space (RKHS) corresponding to the GP kernel, we derive both probabilistic and deterministic bounds on the error of the GPR. We show that these errors are substantially tighter than existing state-of-the-art bounds and are particularly well-suited for GPR with neural network kernels, i.e., Deep Kernel Learning (DKL). Furthermore, motivated by applications in safety-critical domains, we illustrate how these bounds can be combined with stochastic barrier functions to successfully quantify the safety probability of an unknown dynamical system from finite data. We validate the efficacy of our approach through several benchmarks and comparisons against existing bounds. The results show that our bounds are consistently smaller, and that DKLs can produce error bounds tighter than sample noise, significantly improving the safety probability of control systems.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑