Quantile-Based Optimization of Noisy Computer Experiments With Tunable Precision

By Victor Picheny, David Ginsbourger, Yann Richet, Gregory Caplin. 2013

216 citations recorded by the source corpus.

Subfields: Computational Theory and Mathematics, Management Science and Operations Research, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

1. Yaochu Jin
2. Eytan Bakshy
3. Guilherme Ottoni
4. Victor Picheny
5. Peter I. Frazier
6. David Ginsbourger
7. Markus Olhofer
8. Robert B. Gramacy

In Management Science and Operations Research

1. Guilherme Ottoni
2. Yaochu Jin
3. Eytan Bakshy
4. Peter I. Frazier
5. Markus Olhofer
6. Victor Picheny
7. Robert B. Gramacy
8. David Ginsbourger

In Statistics and Probability

1. Robert B. Gramacy
2. Peter I. Frazier
3. Herbert K. H. Lee
4. David Ginsbourger
5. Victor Picheny
6. Yaochu Jin
7. Jialei Wang
8. Matthias Poloczek

Citations in the corpus

• An evaluation of adaptive surrogate modeling based optimization with two benchmark problems (2014, methodology): “For this reason, MEI tends to put more points near the boundary of the feasible space in the initial search stage.”
• Recent Advances in Bayesian Optimization (2022, methodology): “Alternatively, the β-quantile given by the GP surrogate is used as a reference in [153].”
• Multi-Information Source Optimization (2016, methodology): “Subsequently, the expected improvement method was extended to deal with observational noise, e.g., see Huang et al. [2006b], Scott et al. [2011], Picheny et al. [2013].”
• Modeling an augmented Lagrangian for improved blackbox constrained optimization (2014, background): “Assuming normal predictive equations, we simply sample y c 1 ( x ) , . . . , y and N ( µ nc i , σ 2 nc i ), respectively, and then average: We find generally low Monte Carlo error, and hence very few samples (e.g., T = 100) suffice.”
• Constrained Bayesian Optimization with Noisy Experiments (2017, methodology): “Picheny et al. (2013) extend this idea by substituting a GP quantile in the place of the mean, and then optimize expected improvement of that quantile.”
• A benchmark of kriging-based infill criteria for noisy optimization (2013, background): “The variety of applications has resulted in different approaches for Noisy Kriging-based Optimization (NKO) over the last years (Forrester et al. 2006; Huang et al. 2006b; Vazquez et al. 2008; Osborne et al. 2009; Picheny et al. 2013; Srinivas et al. 2010; Scott et al. 2011).”