Probabilistic bisection with spatial metamodels
From MaRDI portal
Publication:2184151
DOI10.1016/J.EJOR.2020.03.049zbMATH Open1443.62066arXiv1807.00095OpenAlexW3013983699MaRDI QIDQ2184151FDOQ2184151
Publication date: 27 May 2020
Published in: European Journal of Operational Research (Search for Journal in Brave)
Abstract: Probabilistic Bisection Algorithm performs root finding based on knowledge acquired from noisy oracle responses. We consider the generalized PBA setting (G-PBA) where the statistical distribution of the oracle is unknown and location-dependent, so that model inference and Bayesian knowledge updating must be performed simultaneously. To this end, we propose to leverage the spatial structure of a typical oracle by constructing a statistical surrogate for the underlying logistic regression step. We investigate several non-parametric surrogates, including Binomial Gaussian Processes (B-GP), Polynomial, Kernel, and Spline Logistic Regression. In parallel, we develop sampling policies that adaptively balance learning the oracle distribution and learning the root. One of our proposals mimics active learning with B-GPs and provides a novel look-ahead predictive variance formula. The resulting gains of our Spatial PBA algorithm relative to earlier G-PBA models are illustrated with synthetic examples and a challenging stochastic root finding problem from Bermudan option pricing.
Full work available at URL: https://arxiv.org/abs/1807.00095
Recommendations
- Probabilistic bisection converges almost as quickly as stochastic approximation
- Bisection search with noisy responses
- A novel technique for stochastic root-finding: enhancing the search with adaptive \(d\)-ary search
- Adaptive designs for stochastic root-finding
- A Bayesian stochastic approximation method
Bayesian inference (62F15) Derivative securities (option pricing, hedging, etc.) (91G20) Sequential statistical design (62L05) Nonconvex programming, global optimization (90C26)
Cites Work
- Efficient global optimization of expensive black-box functions
- The elements of statistical learning. Data mining, inference, and prediction
- Bayesian data analysis.
- Elements of Information Theory
- Stochastic Kriging for Simulation Metamodeling
- Title not available (Why is that?)
- Sequential design of computer experiments for the estimation of a probability of failure
- An informational approach to the global optimization of expensive-to-evaluate functions
- Quantifying Uncertainties on Excursion Sets Under a Gaussian Random Field Prior
- On Deriving the Inverse of a Sum of Matrices
- Title not available (Why is that?)
- A method for the updating of stochastic Kriging metamodels
- Title not available (Why is that?)
- Lipschitzian optimization without the Lipschitz constant
- Maximum Entropy Sampling and Optimal Bayesian Experimental Design
- The stochastic root-finding problem
- Advanced Lectures on Machine Learning
- The Correlated Knowledge Gradient for Simulation Optimization of Continuous Parameters using Gaussian Process Regression
- Twenty Questions with Noise: Bayes Optimal Policies for Entropy Loss
- Probabilistic Bisection Converges Almost as Quickly as Stochastic Approximation
- An information-theoretic analysis of Thompson sampling
- Sequential design strategies for mean response surface metamodeling via stochastic kriging with adaptive exploration and exploitation
- Comparison of Kriging-based algorithms for simulation optimization with heterogeneous noise
- Bisection search with noisy responses
- Constrained Bayesian optimization with noisy experiments
Uses Software
This page was built for publication: Probabilistic bisection with spatial metamodels
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2184151)
