每年出版 6 期
ISSN 打印: 2152-5080
ISSN 在线: 2152-5099
Indexed in
PHYSICS-BASED COVARIANCE MODELS FOR GAUSSIAN PROCESSES WITH MULTIPLE OUTPUTS
摘要
Gaussian process analysis of processes with multiple outputs is limited by the fact that far fewer good classes of covariance functions exist compared with the scalar (single-output) case. To address this difficulty, we turn to covariance function models that take a form consistent in some sense with physical laws that govern the underlying simulated process. Models that incorporate such information are suitable when performing uncertainty quantification or inferences on multidimensional processes with partially known relationships among different variables, also known as cokriging. One example is in atmospheric dynamics where pressure and wind speed are driven by geostrophic assumptions (wind ∝ ∂/∂x pressure). In this study we develop both analytical and numerical auto-covariance and cross-covariance models that are consistent with physical constraints or can incorporate automatically sensible assumptions about the process that generated the data. We also determine high-order closures, which are required for nonlinear dependencies among the observables. We use these models to study Gaussian process regression for processes with multiple outputs and latent processes (i.e., processes that are not directly observed and predicted but inter-relate the output quantities). Our results demonstrate the effectiveness of the approach on both synthetic and real data sets.
-
Kleijnen Jack P.C., Mehdad Ehsan, Multivariate versus univariate Kriging metamodels for multi-response simulation models, European Journal of Operational Research, 236, 2, 2014. Crossref
-
Mahmood Asif, Wolpert Robert L., Pitman E. Bruce, A Physics-Based Emulator for the Simulation of Geophysical Mass Flows, SIAM/ASA Journal on Uncertainty Quantification, 3, 1, 2015. Crossref
-
Fahrenholtz Samuel J., Moon Tim Y., Franco Michael, Medina David, Danish Shabbar, Gowda Ashok, Shetty Anil, Maier Florian, Hazle John D., Stafford Roger J., Warburton Tim, Fuentes David, A model evaluation study for treatment planning of laser-induced thermal therapy, International Journal of Hyperthermia, 31, 7, 2015. Crossref
-
Chiplunkar Ankit, Rachelson Emmanuel, Colombo Michele, Morlier Joseph, Adding Flight Mechanics to Flight Loads Surrogate Model using Multi-Output Gaussian Processes, 17th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference, 2016. Crossref
-
Chiplunkar Ankit, Rachelson Emmanuel, Colombo Michele, Morlier Joseph, Approximate Inference in Related Multi-output Gaussian Process Regression, in Pattern Recognition Applications and Methods, 10163, 2017. Crossref
-
Yung Joshua P., Fuentes David, MacLellan Christopher J, Maier Florian, Liapis Yannis, Hazle John D., Stafford R. Jason, Referenceless magnetic resonance temperature imaging using Gaussian process modeling, Medical Physics, 44, 7, 2017. Crossref
-
Fan Minjie, Paul Debashis, Lee Thomas C. M., Matsuo Tomoko, Modeling Tangential Vector Fields on a Sphere, Journal of the American Statistical Association, 113, 524, 2018. Crossref
-
Yoo Byunghyun, Kim Jinwhan, Probabilistic modeling of ship powering performance using full-scale operational data, Applied Ocean Research, 82, 2019. Crossref
-
Kleijnen Jack P. C., Mehdad Ehsan, Kriging in Multi-Response Simulation, Including a Monte Carlo Laboratory, SSRN Electronic Journal, 2012. Crossref
-
Wu Jin-Long, Michelén-Ströfer Carlos, Xiao Heng, Physics-informed covariance kernel for model-form uncertainty quantification with application to turbulent flows, Computers & Fluids, 193, 2019. Crossref
-
Chmiela Stefan, Sauceda Huziel E., Tkatchenko Alexandre, Müller Klaus-Robert, Accurate Molecular Dynamics Enabled by Efficient Physically Constrained Machine Learning Approaches, in Machine Learning Meets Quantum Physics, 968, 2020. Crossref
-
Serafino Aldo, Obert Benoit, Cinnella Paola, Multi-Fidelity Gradient-Based Strategy for Robust Optimization in Computational Fluid Dynamics, Algorithms, 13, 10, 2020. Crossref
-
Bessac Julie, Constantinescu Emil, Anitescu Mihai, Stochastic simulation of predictive space–time scenarios of wind speed using observations and physical model outputs, The Annals of Applied Statistics, 12, 1, 2018. Crossref
-
Sahastrabuddhe Rishi, Ghosh Subimal, Does statistical model perform at par with computationally expensive general circulation model for decadal prediction?, Environmental Research Letters, 16, 6, 2021. Crossref
-
Genton Marc G., Kleiber William, Cross-Covariance Functions for Multivariate Geostatistics, Statistical Science, 30, 2, 2015. Crossref
-
Hanuka Adi, Huang X., Shtalenkova J., Kennedy D., Edelen A., Zhang Z., Lalchand V. R., Ratner D., Duris J., Physics model-informed Gaussian process for online optimization of particle accelerators, Physical Review Accelerators and Beams, 24, 7, 2021. Crossref
-
Peli Riccardo, Menafoglio Alessandra, Cervino Marianna, Dovera Laura, Secchi Piercesare, Physics-based Residual Kriging for dynamically evolving functional random fields, Stochastic Environmental Research and Risk Assessment, 2022. Crossref
-
Cortes Cristian L., Lefebvre Pascal, Lauk Nikolai, Davis Michael J., Sinclair Neil, Gray Stephen K., Oblak Daniel, Sample-Efficient Adaptive Calibration of Quantum Networks Using Bayesian Optimization, Physical Review Applied, 17, 3, 2022. Crossref
-
Hebbal Ali, Balesdent Mathieu, Brevault Loïc, Melab Nouredine, Talbi El-Ghazali, Deep Gaussian process for multi-objective Bayesian optimization, Optimization and Engineering, 2022. Crossref
-
Marinescu Marius, Olivares Alberto, Staffetti Ernesto, Sun Junzi, Chen Chi-Hua, Wind velocity field estimation from aircraft derived data using Gaussian process regression, PLOS ONE, 17, 10, 2022. Crossref