Affiliation:
1. Columbia University, USA
2. Tongji University, China
3. Carnegie Mellon University, USA
Abstract
We study rare-event simulation for a class of problems where the target hitting sets of interest are defined via modern machine learning tools such as neural networks and random forests. This problem is motivated from fast emerging studies on the safety evaluation of intelligent systems, robustness quantification of learning models, and other potential applications to large-scale simulation in which machine learning tools can be used to approximate complex rare-event set boundaries. We investigate an importance sampling scheme that integrates the dominating point machinery in large deviations and sequential mixed integer programming to locate the underlying dominating points. Our approach works for a range of neural network architectures including fully connected layers, rectified linear units, normalization, pooling and convolutional layers, and random forests built from standard decision trees. We provide efficiency guarantees and numerical demonstration of our approach using a classification model in the UCI Machine Learning Repository.
Funder
National Science Foundation
Manufacturing Futures Initiative at Carnegie Mellon University
Publisher
Association for Computing Machinery (ACM)
Subject
Computer Science Applications,Modeling and Simulation
Reference122 articles.
1. Efficient Monte Carlo for high excursions of Gaussian random fields
2. Adaptive Importance Sampling Technique for Markov Chains Using Stochastic Approximation
3. Efficient Simulation for Expectations over the Union of Half-Spaces
4. Mansur Arief, Zhiyuan Huang, Guru Koushik Senthil Kumar, Yuanlu Bai, Shengyi He, Wenhao Ding, Henry Lam, and Ding Zhao. 2021. Deep probabilistic accelerated evaluation: A robust certifiable rare-event simulation methodology for black-box safety-critical systems. In International Conference on Artificial Intelligence and Statistics. PMLR, 595–603.
5. Conjugate processes and the silumation of ruin problems
Cited by
9 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献