Batch-sequential design and heteroskedastic surrogate modeling for delta smelt conservation
From MaRDI portal
Publication:2154181
DOI10.1214/21-AOAS1521zbMATH Open1498.62298arXiv2010.06515OpenAlexW3093417780MaRDI QIDQ2154181FDOQ2154181
Authors: Yanyan Li 
Publication date: 14 July 2022
Published in: The Annals of Applied Statistics (Search for Journal in Brave)
Abstract: Delta smelt is an endangered fish species in the San Francisco estuary that have shown an overall population decline over the past 30 years. Researchers have developed a stochastic, agent-based simulator to virtualize the system, with the goal of understanding the relative contribution of natural and anthropogenic factors suggested as playing a role in their decline. However, the input configuration space is high-dimensional, running the simulator is time-consuming, and its noisy outputs change nonlinearly in both mean and variance. Getting enough runs to effectively learn input--output dynamics requires both a nimble modeling strategy and parallel supercomputer evaluation. Recent advances in heteroskedastic Gaussian process (HetGP) surrogate modeling helps, but little is known about how to appropriately plan experiments for highly distributed simulator evaluation. We propose a batch sequential design scheme, generalizing one-at-a-time variance-based active learning for HetGP surrogates, as a means of keeping multi-core cluster nodes fully engaged with expensive runs. Our acquisition strategy is carefully engineered to favor selection of replicates which boost statistical and computational efficiencies when training surrogates to isolate signal in high noise regions. Design and modeling performance is illustrated on a range of toy examples before embarking on a large-scale smelt simulation campaign and downstream high-fidelity input sensitivity analysis.
Full work available at URL: https://arxiv.org/abs/2010.06515
Recommendations
- Sequential design for computer experiments with a flexible Bayesian additive model
- Design and analysis of simulation experiments
- A novel hybrid sequential design strategy for global surrogate modeling of computer experiments
- Bayesian-validated computer-simulation surrogates for optimization and design: Error estimates and applications
- Optimal design for correlated processes with input-dependent noise
sensitivity analysisinput-dependent noisereplicationagent-based modelactive learningGaussian process surrogate modeling
Bayesian inference (62F15) Sequential statistical design (62L05) Applications of statistics to environmental and related topics (62P12)
Cites Work
- Efficient global optimization of expensive black-box functions
- Discrete Optimization via Simulation Using COMPASS
- Strictly Proper Scoring Rules, Prediction, and Estimation
- Design and analysis of computer experiments. With comments and a rejoinder by the authors
- Bayesian calibration of computer models. (With discussion)
- Stochastic Kriging for Simulation Metamodeling
- A Limited Memory Algorithm for Bound Constrained Optimization
- Practical Heteroscedastic Gaussian Process Modeling for Large Simulation Experiments
- Microcolony and biofilm formation as a survival strategy for bacteria
- Optimal predictive model selection.
- Locally induced Gaussian processes for large-scale simulation experiments
- Analyzing stochastic computer models: a review with opportunities
- A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code
- Microsimulation Model Calibration using Incremental Mixture Approximate Bayesian Computation
- Title not available (Why is that?)
- Probabilistic Sensitivity Analysis of Complex Models: A Bayesian Approach
- The design and analysis of computer experiments
- Exploratory designs for computational experiments
- Batch sequential designs for computer experiments
- Title not available (Why is that?)
- Calibrating a Stochastic, Agent-Based Model Using Quantile-Based Emulation
- Computer experiment designs for accurate prediction
- Evaluating Gaussian process metamodels and sequential designs for noisy level set estimation
- An efficient algorithm for Elastic I‐optimal design of generalized linear models
- Sequential Learning of Active Subspaces
Uses Software
This page was built for publication: Batch-sequential design and heteroskedastic surrogate modeling for delta smelt conservation
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2154181)
