Refine
Year of publication
Document Type
- Article (14)
- In Proceedings (7)
- ZIB-Report (6)
- Bachelor's Thesis (1)
- Book chapter (1)
- In Collection (1)
- Research data (1)
Is part of the Bibliography
- no (31)
Keywords
- importance sampling (2)
- metadynamics (2)
- metastability (2)
- neural networks (2)
- rare event simulation (2)
- stochastic optimal control (2)
- Donsker-Varadhan principle (1)
- Green's functions (1)
- Importance sampling (1)
- active thermography (1)
Institute
- Numerical Mathematics (15)
- Modeling and Simulation of Complex Processes (12)
- Computational Nano Optics (8)
- Computational Medicine (3)
- Visual Data Analysis (2)
- Visual and Data-centric Computing (2)
- ZIB Allgemein (2)
- Computational Molecular Design (1)
- Image Analysis in Biology and Materials Science (1)
- Mathematical Optimization (1)
- Mathematics for Life and Materials Science (1)
Importance sampling is a popular variance reduction method for Monte Carlo estimation, where an evident question is how to design good proposal distributions. While in most cases optimal (zero-variance) estimators are theoretically possible, in practice only suboptimal proposal distributions are available and it can often be observed numerically that those can reduce statistical performance significantly, leading to large relative errors and therefore counteracting the original intention. Previous analysis on importance sampling has often focused on asymptotic arguments that work well in a large deviations regime. In this article, we provide lower and upper bounds on the relative error in a nonasymptotic setting. They depend on the deviation of the actual proposal from optimality, and we thus identify potential robustness issues that importance sampling may have, especially in high dimensions. We particularly focus on path sampling problems for diffusion processes with nonvanishing noise, for which generating good proposals comes with additional technical challenges. We provide numerous numerical examples that support our findings and demonstrate the applicability of the derived bounds.