Dario Azzimonti presents a method for estimating orthant probabilities of high-dimensional Gaussian vectors. The procedure combines deterministic quadrature for low-dimensional terms with a novel asymmetric nested Monte Carlo algorithm for the remainder. The method is applied to derive conservative estimates of excursion sets for expensive deterministic functions under a Gaussian random field prior.
Use Cases
- Benchmarking probability estimation algorithms based on the described numerical study.
- Estimating excursion sets for expensive deterministic functions based on the Gaussian random field prior application.
- Comparing stochastic and deterministic techniques for high-dimensional probability computation based on the two-step procedure.
Strengths
- Method is compared against state-of-the-art techniques in a numerical study.
- Procedure highlights cases where the asymmetric nested Monte Carlo approximation brings substantial efficiency gains.
Limitations
- Description metadata is limited; actual data quality requires manual inspection after download.
- Column-level documentation is absent; field semantics must be inferred after download.
Provenance
- Source
- Dario Azzimonti