Multi-Resolution Functional ANOVA for Large-Scale, Many-Input Computer Experiments

DOI10.1080/01621459.2019.1595630zbMATH Open1445.62189arXiv1709.07064OpenAlexW2962864460WikidataQ128184903 ScholiaQ128184903MaRDI QIDQ147162FDOQ147162

Matthew Plumlee, Wenjia Wang, Chih-Li Sung, Chih-Li Sung, Benjamin Haaland, Benjamin Haaland, Matthew Plumlee, Wenjia Wang

Publication date: 20 September 2017

Published in: Journal of the American Statistical Association (Search for Journal in Brave)

Abstract: The Gaussian process is a standard tool for building emulators for both deterministic and stochastic computer experiments. However, application of Gaussian process models is greatly limited in practice, particularly for large-scale and many-input computer experiments that have become typical. We propose a multi-resolution functional ANOVA model as a computationally feasible emulation alternative. More generally, this model can be used for large-scale and many-input non-linear regression problems. An overlapping group lasso approach is used for estimation, ensuring computational feasibility in a large-scale and many-input setting. New results on consistency and inference for the (potentially overlapping) group lasso in a high-dimensional setting are developed and applied to the proposed multi-resolution functional ANOVA model. Importantly, these results allow us to quantify the uncertainty in our predictions. Numerical examples demonstrate that the proposed model enjoys marked computational advantages. Data capabilities, both in terms of sample size and dimension, meet or exceed best available emulation tools while meeting or exceeding emulation accuracy.

Full work available at URL: https://arxiv.org/abs/1709.07064

zbMATH Keywords

computer experimentsnonlinear regressionlarge-scalemany-inputoverlapping group Lasso

Mathematics Subject Classification ID

Statistical aspects of big data and data science (62R07) Analysis of variance and covariance (ANOVA) (62J10) Computational aspects of data analysis and big data (68T09) Functional data analysis (62R10)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
• Strictly Proper Scoring Rules, Prediction, and Estimation
• Bagging predictors
• Lasso-type recovery of sparse representations for high-dimensional data
• Multivariate adaptive regression splines
• Smoothing noisy data with spline functions: Estimating the correct degree of smoothing by the method of generalized cross-validation
• The design and analysis of computer experiments.
• Design and analysis of computer experiments. With comments and a rejoinder by the authors
• Piecewise polynomial, positive definite and compactly supported radial functions of minimal degree
• Model Selection and Estimation in Regression with Grouped Variables
• Theory of Reproducing Kernels
• On bagging and nonlinear estimation
• Smoothing spline ANOVA for exponential families, with application to the Wisconsin epidemiological study of diabetic retinopathy. (The 1994 Neyman Memorial Lecture)
• The Group Lasso for Logistic Regression
• Regularization Parameter Selections via Generalized Information Criterion
• Tuning parameter selectors for the smoothly clipped absolute deviation method
• Efficient emulators of computer experiments using compactly supported correlation functions, with an application to cosmology
• Analyzing bagging
• Stochastic simulation: Algorithms and analysis
• Approximate efficiency of a selection procedure for the number of regression variables
• Sparse Additive Models
• Fast Prediction of Deterministic Functions Using Sparse Grid Experimental Designs
• Component selection and smoothing in multivariate nonparametric regression
• Asymptotic optimality for \(C_ p\), \(C_ L\), cross-validation and generalized cross-validation: Discrete index set
• Polynomial splines and their tensor products in extended linear modeling. (With discussions)
• Functional aggregation for nonparametric regression.
• Stochastic processes with sample paths in reproducing kernel Hilbert spaces
• Massively Parallel Approximate Gaussian Process Regression
• Monte Carlo Variance of Scrambled Net Quadrature
• Accurate emulators for large-scale computer experiments
• Exploiting Variance Reduction Potential in Local Gaussian Process Search
• A general theory of hypothesis tests and confidence regions for sparse high dimensional models
• Scattered Data Approximation

Cited In (8)

• Projection pursuit emulation for many-input computer experiments
• Sequential Learning of Active Subspaces
• A Clustered Gaussian Process Model for Computer Experiments
• Boosting on the functional ANOVA decomposition
• On the inference of applying Gaussian process modeling to a deterministic function
• Predictive Subdata Selection for Computer Models
• MRFA
• Linear screening for high-dimensional computer experiments

Uses Software

• R
• gss
• laGP
• mlegp
• GPfit

This page was built for publication: Multi-Resolution Functional ANOVA for Large-Scale, Many-Input Computer Experiments

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q147162)