使用状态依赖重要性采样提议估计黑盒自主系统的失效概率
Failure Probability Estimation for Black-Box Autonomous Systems using State-Dependent Importance Sampling Proposals
AI总结：
针对黑盒安全关键自主系统罕见失效难以估计的问题，提出基于状态依赖提议和Markov得分上升的自适应重要性采样算法，在四个序贯系统上显著提升估计精度并开源。
摘要：
估计失效概率是开发安全关键型自主系统的关键步骤。由于此类系统中的失效十分罕见，Monte Carlo采样等直接估计方法往往不切实际。现有重要性采样方法无法扩展到具有大状态空间和长时域的序贯决策系统。为解决这些局限，我们提出一种自适应重要性采样算法。该方法最小化状态依赖提议分布与最优重要性采样分布的松弛形式之间的前向Kullback-Leibler散度，并使用Markov得分上升方法估计该目标。我们在四个序贯系统上评估了所提方法，结果表明，相比基线Monte Carlo和重要性采样技术，它能提供更准确的失效概率估计。本工作已开源。
英文摘要：
Estimating the probability of failure is a critical step in developing safety-critical autonomous systems. Direct estimation methods such as Monte Carlo sampling are often impractical due to the rarity of failures in these systems. Existing importance sampling approaches do not scale to sequential decision-making systems with large state spaces and long horizons. We propose an adaptive importance sampling algorithm to address these limitations. Our method minimizes the forward Kullback-Leibler divergence between a state-dependent proposal distribution and a relaxed form of the optimal importance sampling distribution. Our method uses Markov score ascent methods to estimate this objective. We evaluate our approach on four sequential systems and show that it provides more accurate failure probability estimates than baseline Monte Carlo and importance sampling techniques. This work is open sourced.
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。
