Article
Engineering, Civil
Dayang Li, Lucy Marshall, Zhongmin Liang, Ashish Sharma
Summary: By applying the Variational Bayesian Long Short-Term Memory network (VB-LSTM) approach to hydrological models, it can improve the accuracy of deterministic and probabilistic predictions, show better robustness, and be less impacted by the selection of ensemble members.
JOURNAL OF HYDROLOGY
(2022)
Article
Engineering, Mechanical
Wyatt Bridgman, Reese E. Jones, Mohammad Khalil
Summary: This paper proposes a method for improving variational inference by constructing an initial Gaussian mixture model approximation, and it is demonstrated through synthetic tests and inversion problems in structural dynamics.
PROBABILISTIC ENGINEERING MECHANICS
(2023)
Article
Engineering, Multidisciplinary
Hening Huang
Summary: This paper proposes a propensity-based framework for measurement uncertainty analysis. The measurand is regarded as a random variable characterized by central tendency and dispersion. The state of propensity is described by a probability density function (PDF). The framework encodes the state of propensity of the measurand based on all available information about influence quantities.
Article
Engineering, Geological
Guilherme J. C. Gomes, John H. Forero, Euripedes A. Vargas, Jasper A. Vrugt
Summary: This paper discusses the importance of understanding rock anisotropy for reliability analysis and engineering design, and proposes a formulation for calculating strength anisotropy embedded in a Bayesian framework for practical application. The results demonstrate the accuracy of the proposed model in predicting peak strengths of rocks with varying degrees of anisotropy, confining stresses and anisotropy orientations, emphasizing the need for explicit treatment of strength anisotropy uncertainty in rock mechanics studies. The Bayesian methodology is versatile and can assist in informing geotechnical engineers, contractors, and other professionals about rock conditions, design reliability, and overall risks of engineering structures.
INTERNATIONAL JOURNAL OF ROCK MECHANICS AND MINING SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Derek S. Prijatelj, Mel McCurrie, Samuel E. Anthony, Walter J. Scheirer
Summary: An interesting development in automatic visual recognition is the emergence of tasks where objective labels cannot be assigned to images, but human judgements can still be collected. This study proposes a Bayesian framework for evaluating black box predictors in this scenario, providing a method for estimating the epistemic uncertainty of the predictors. The framework is successfully applied to four image classification tasks that use subjective human judgements.
PATTERN RECOGNITION
(2022)
Article
Materials Science, Multidisciplinary
Arun Hegde, Elan Weiss, Wolfgang Windl, Habib Najm, Cosmin Safta
Summary: Developing reliable interatomic potential models is crucial for atomistic simulations. This study investigates the possibility of fitting EAM potentials for binary alloys using Bayesian calibration. By using predictive distributions, the limitations of the potential are demonstrated, emphasizing the importance of statistical formulations for model error.
COMPUTATIONAL MATERIALS SCIENCE
(2022)
Article
Meteorology & Atmospheric Sciences
Melanie Bieli, Oliver R. A. Dunbar, Emily K. de Jong, Anna Jaruga, Tapio Schneider, Tobias Bischoff
Summary: This article introduces a new bulk microphysics scheme called Cloudy and demonstrates how Bayesian learning can be applied to infer parameters. By using the CES algorithm, computational efficiency is improved and noise pollution is reduced, leading to successful results in experiments.
JOURNAL OF ADVANCES IN MODELING EARTH SYSTEMS
(2022)
Article
Robotics
Jasprabhjit Mehami, Raphael Falque, Teresa Vidal-Calleja, Alen Alempijevic
Summary: This study proposes a physics-based, data-driven light-source-mean model for reflectance estimation, achieving improved accuracy by utilizing multi-modal sensing information and shape information obtained by depth cameras. Experimental results demonstrate that the proposed model outperforms existing methods, reducing the error by 96.8% on average. The improved reflectance estimation method is further validated through a multi-modal classification application.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Environmental Sciences
Pedram Darbandsari, Paulin Coulibaly
Summary: This study evaluates the impact of different hydrologic models on the performance of the hydrologic uncertainty processor (HUP) and proposes a multimodel Bayesian postprocessor (HUP-BMA). Results demonstrate the superiority of HUP-BMA in quantifying hydrologic uncertainty and forecasting compared to traditional HUP and BMA methods.
WATER RESOURCES RESEARCH
(2021)
Article
Engineering, Biomedical
Selma Metzner, Gerd Wubbeler, Sebastian Flassbeck, Constance Gatefait, Christoph Kolbitsch, Clemens Elster
Summary: Magnetic Resonance Fingerprinting (MRF) is a promising technique for fast quantitative imaging of human tissue, providing valuable diagnostic parameters like T-1 and T-2 MR relaxation times. A Bayesian approach is proposed for uncertainty quantification of dictionary-based MRF, leading to probability distributions for T-1 and T-2 in every voxel. The method successfully characterizes uncertainties in relaxation time estimates and is consistent with observed variability in simulations and in vivo measurements.
PHYSICS IN MEDICINE AND BIOLOGY
(2021)
Article
Environmental Sciences
K. C. Abbaspour
Summary: Using parameters from the best-fit simulation to represent a calibrated hydrological model may lead to misleading results, as the best solution's parameters may differ significantly from the next best set. The non-uniqueness of objective function values in calibration poses challenges in interpreting watershed processes accurately, and researchers must consider model output uncertainty to assess calibration/validation effectively.
SCIENCE OF THE TOTAL ENVIRONMENT
(2022)
Article
Computer Science, Interdisciplinary Applications
Nikolaos M. Dimitriou, Ece Demirag, Katerina Strati, Georgios D. Mitsis
Summary: This study investigates the performance of mathematical models of tumour growth through integrating multiscale measurements. The results demonstrate that incorporating multiscale measurements can improve model performance, particularly when high-dose treatment is involved.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2024)
Article
Computer Science, Interdisciplinary Applications
Adam P. Generale, Surya R. Kalidindi
Summary: This paper presents a probabilistic framework for improving microstructure-sensitive predictions of multiscale modeling of heterogeneous materials. The framework calibrates complex constitutive models at the mesoscale to sparsely observed macroscale experimental data and propagates the stochastic constituent behavior at the mesoscale to low-cost homogenized predictions. The methodology can be widely applied to various material classes and constitutive models with high-dimensional parameter sets.
COMPUTERS & STRUCTURES
(2023)
Article
Engineering, Mechanical
Patricio Peralta, O. Rafael Ruiz, Hussein Rappel, P. A. Stephane Bordas
Summary: This new framework utilizes dynamic estimators to infer the electromechanical properties in Piezoelectric Energy Harvesters (PEHs), overcoming the mismatch issue in updating properties associated with a set of PEHs. By modifying the likelihood function, the framework is able to account for a predictive model with three outputs obtained from the FRF.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Construction & Building Technology
Lin Lin, Guodong Chen, Xiaochen Liu, Xiaohua Liu, Tao Zhang
Summary: This paper investigates the cooling load characteristics in airport terminal buildings and categorizes the indoor areas based on their distinct characteristics. The study reveals the significant impact of outdoor temperature, outdoor relative humidity, and air change rate on cooling loads. The Bayesian calibration analysis identifies the differences in main parameters as the key factors causing the discrepancy between measurement and design. The paper also discusses the potential for flexible cooling load adjustment across different area categories.
JOURNAL OF BUILDING ENGINEERING
(2023)
Article
Computer Science, Interdisciplinary Applications
Roya A. Cody, Bryan A. Tolson, Jeff Orchard
JOURNAL OF COMPUTING IN CIVIL ENGINEERING
(2020)
Article
Water Resources
Ming Han, Juliane Mai, Bryan A. Tolson, James R. Craig, Etienne Gaborit, Hongli Liu, Konhee Lee
CANADIAN WATER RESOURCES JOURNAL
(2020)
Article
Computer Science, Interdisciplinary Applications
James R. Craig, Genevieve Brown, Robert Chlumsky, R. Wayne Jenkinson, Georg Jost, Konhee Lee, Juliane Mai, Martin Serrer, Nicholas Sgro, Mahyar Shafii, Andrew P. Snowdon, Bryan A. Tolson
ENVIRONMENTAL MODELLING & SOFTWARE
(2020)
Article
Environmental Sciences
Diana Spieler, Juliane Mai, James R. Craig, Bryan A. Tolson, Niels Schuetze
WATER RESOURCES RESEARCH
(2020)
Article
Environmental Sciences
Juliane Mai, Richard Arsenault, Bryan A. Tolson, Marco Latraverse, Kenjy Demeester
WATER RESOURCES RESEARCH
(2020)
Article
Environmental Sciences
Robert Chlumsky, Juliane Mai, James R. Craig, Bryan A. Tolson
Summary: The improvement of hydrological modeling frameworks allows for both model structure and parameters to be automatically calibrated and evaluated. The blended model structure calibration method can identify near-optimal model structures at significantly lower computational cost, as well as help identify dominant processes and model structures in catchments.
WATER RESOURCES RESEARCH
(2021)
Article
Engineering, Civil
Juliane Mai, Bryan A. Tolson, Hongren Shen, Etienne Gaborit, Vincent Fortin, Nicolas Gasset, Herve Awoye, Tricia A. Stadnyk, Lauren M. Fry, Emily A. Bradley, Frank Seglenieks, Andre G. T. Temgoua, Daniel G. Princz, Shervan Gharari, Amin Haghnegahdar, Mohamed E. Elshamy, Saman Razavi, Martin Gauch, Jimmy Lin, Xiaojing Ni, Yongping Yuan, Meghan McLeod, Nandita B. Basu, Rohini Kumar, Oldrich Rakovec, Luis Samaniego, Sabine Attinger, Narayan K. Shrestha, Prasad Daggupati, Tirthankar Roy, Sungwook Wi, Tim Hunter, James R. Craig, Alain Pietroniro
Summary: This study evaluates the performance of hydrologic models in simulating variables in the Lake Erie watershed, finding that Machine Learning models perform well in calibration but decrease in validation, models calibrated at individual stations perform equally well, and most distributed models have trouble simulating urban areas but excel in validation.
JOURNAL OF HYDROLOGIC ENGINEERING
(2021)
Article
Computer Science, Interdisciplinary Applications
Juliane Mai, James R. Craig, Bryan A. Tolson
Summary: This article provides a simple algorithm for randomly sampling a set of weights with their sum constrained to be equal to one. The algorithm has potential applications in calibration, uncertainty analysis, and sensitivity analysis of environmental models. The author demonstrates the efficiency and superiority of the proposed method compared to alternative sampling methods through three example applications.
ENVIRONMENTAL MODELLING & SOFTWARE
(2022)
Article
Water Resources
Hongli Liu, Bryan A. Tolson, Andrew J. Newman, Andrew W. Wood
Summary: This study examines the potential of using a forcing ensemble to identify robust parameters through model calibration, comparing ensemble forcing-based calibration with deterministic forcing-based calibration. Results show that ensemble calibration generates less biased parameter estimates and improves the overall reliability and simulation skill of ensemble simulation results by reducing the risk of inaccurate flow simulation caused by poor-quality meteorological inputs.
HYDROLOGICAL PROCESSES
(2021)
Article
Environmental Sciences
Hongren Shen, Bryan A. Tolson, Juliane Mai
Summary: This study empirically assesses how different data splitting methods influence post-validation model testing period performance in hydrological modeling. The findings suggest that calibrating to older data and then validating models on newer data produces inferior model testing period performance, while calibrating to the full available data and skipping model validation is the most robust split-sample decision. The experimental findings remain consistent across different factors and strongly support revising the traditional split-sample approach in hydrological modeling.
WATER RESOURCES RESEARCH
(2022)
Article
Multidisciplinary Sciences
Juliane Mai, James R. Craig, Bryan A. Tolson, Richard Arsenault
Summary: The sensitivity of streamflow simulations to different hydrologic processes is analyzed in this study, using a novel analysis method that considers both parametric and structural uncertainties. The results show that quickflow is the most sensitive process for streamflow simulations across North America. Approximations of model process and parameter sensitivities are derived based on physiographic and climatologic data, and detailed spatio-temporal inputs and results are shared through an interactive website.
NATURE COMMUNICATIONS
(2022)
Review
Computer Science, Interdisciplinary Applications
Marjan Asgari, Wanhong Yang, John Lindsay, Bryan Tolson, Maryam Mehri Dehnavi
Summary: This paper reviews the application of parallel computing in calibrating watershed hydrologic models and summarizes their contributions, knowledge gaps, and future research directions. The studies parallelized models using random-sampling-based algorithms or optimization algorithms and achieved significant speedup gain and efficiency. However, the speedup gain and efficiency decrease as the number of parallel processing units increases, especially after a certain threshold. Various combinations of hydrologic models, optimization algorithms, parallelization strategies, architectures, and communication modes need to be explored to improve speedup gain, efficiency, and solution quality. A standardized set of performance evaluation metrics should be developed to assess parallelization approaches. Interactive multiobjective optimization algorithms and integrated sensitivity analysis and calibration algorithms can also be potential future research areas.
ENVIRONMENTAL MODELLING & SOFTWARE
(2022)
Article
Environmental Sciences
Martin Gauch, Frederik Kratzert, Oren Gilon, Hoshin Gupta, Juliane Mai, Grey Nearing, Bryan Tolson, Sepp Hochreiter, Daniel Klotz
Summary: Building accurate rainfall-runoff models is crucial in hydrological science and practice. In this study, expert opinions were compared with quantitative metrics, and it was found that experts generally agreed with the metrics and showed a preference for Machine Learning models over traditional hydrological models. Although there were inconsistencies in expert opinions, where there was agreement, the opinions could be predicted from the quantitative metrics.
WATER RESOURCES RESEARCH
(2023)
Article
Geosciences, Multidisciplinary
Juliane Mai, Hongren Shen, Bryan A. Tolson, Etienne Gaborit, Richard Arsenault, James R. Craig, Vincent Fortin, Lauren M. Fry, Martin Gauch, Daniel Klotz, Frederik Kratzert, Nicole O'Brien, Daniel G. Princz, Sinan Rasiya Koya, Tirthankar Roy, Frank Seglenieks, Narayan K. Shrestha, Andre G. T. Temgoua, Vincent Vionnet, Jonathan W. Waddell
Summary: This study conducted a model intercomparison to compare different model setups in simulating outputs in the Great Lakes region. The results showed that the machine-learning-based model performed the best in simulating streamflow, while the locally calibrated models and regionally calibrated models showed varying performances in different areas. The study also compared additional model outputs, such as evapotranspiration, soil moisture, and snow water equivalent, against gridded reference datasets.
HYDROLOGY AND EARTH SYSTEM SCIENCES
(2022)
Article
Geosciences, Multidisciplinary
Juliane Mai, James R. Craig, Bryan A. Tolson
HYDROLOGY AND EARTH SYSTEM SCIENCES
(2020)