Article
Acoustics
Sandip Chajjed, Mohammad Khalil, Dominique Poirel, Chris Pettit, Abhijit Sarkar
Summary: This paper reports the generalization of the Bayesian formulation of the flutter margin method, which improves the predictive performance by incorporating the joint prior of aeroelastic modal parameters. The improved algorithm reduces uncertainties in predicting flutter speed and can cut cost by reducing the number of flight tests.
JOURNAL OF SOUND AND VIBRATION
(2024)
Article
Mathematics
Francisco Javier Diez, Manuel Arias, Jorge Perez-Martin, Manuel Luque
Summary: OpenMarkov is an open-source software tool designed for probabilistic graphical models, primarily in medicine but also used in other fields and education in over 30 countries. This paper explains how OpenMarkov can be used as a pedagogical tool to teach the main concepts of Bayesian networks and influence diagrams, as well as various inference algorithms.
Article
Environmental Sciences
Hao Deng, Shengfang Zhou, Yong He, Zeduo Lan, Yanhong Zou, Xiancheng Mao
Summary: Numerical modeling is crucial for understanding the dynamics of contaminants transport in groundwater. This paper presents a Bayesian optimization method that efficiently calibrates numerical models of groundwater contaminant transport. The method utilizes a probabilistic surrogate model and an expected improvement acquisition function to improve the efficiency of model calibration.
Article
Environmental Sciences
Claudio M. Pierard, Deborah Bassotto, Florian Meirer, Erik van Sebille
Summary: Most marine plastic pollution originates from land, but once in the ocean, it is difficult to determine its source. Researchers have developed a Bayesian inference framework that combines information about plastic emitted by rivers with simulation techniques, allowing them to calculate the probability that a piece of plastic found at sea came from a specific river source. This framework provides the basis for attributing marine plastic pollution to its source.
FRONTIERS IN MARINE SCIENCE
(2022)
Article
Acoustics
S. K. Lai, Y. T. Zhang, J. Q. Sun
Summary: This study investigates a modified active noise control system using Bayesian inference method, combining FxLMS algorithm and dynamic linear model to enhance low-frequency noise attenuation. The combination of Bayesian approach and DLM aids in raw signal pre-processing and generating reference signals, contributing to noise characteristics determination and feedback to the control system for better performance in time-domain signal control algorithms.
Article
Engineering, Multidisciplinary
Pinghe Ni, Jun Li, Hong Hao, Qiang Han, Xiuli Du
Summary: This study proposes a novel variational Bayesian inference approach to estimate posterior probability distributions of civil engineering structures by using vibration responses, improving the accuracy and efficiency of estimating posterior probability distributions in model updating.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2021)
Article
Multidisciplinary Sciences
Celia C. Beron, Shay Q. Neufeld, Scott W. Linderman, Bernardo L. Sabatini
Summary: In probabilistic and nonstationary environments, mice use internal and external cues to make decisions. The behavior of mice in a task with time-varying reward probabilities is both deterministic and stochastic. Modeling their behavior through equivalent models reveals that mice achieve near-maximal reward rates.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2022)
Article
Engineering, Civil
Xue Jiang, Rui Ma, Yanxin Wang, Wenlong Gu, Wenxi Lu, Jin Na
Summary: This study proposes a new two-stage surrogate-assisted Markov chain Monte Carlo-based Bayesian framework for identifying contaminant source parameters in groundwater. An adaptive update feedback process and a multiobjective feasibility-enhanced particle swarm optimization algorithm are utilized to enhance the accuracy and efficiency of the framework.
JOURNAL OF HYDROLOGY
(2021)
Article
Automation & Control Systems
Patrick Kreitzberg, Oliver Serang
Summary: Various methods exist for computing marginal involving a linear Diophantine constraint on random variables, each with limitations. This study introduces a new approach, the trimmed p-convolution tree, which generalizes the applicability of existing methods and achieves better runtime. Additionally, two different methods for approximating max-convolution are introduced using Cartesian product trees.
JOURNAL OF MACHINE LEARNING RESEARCH
(2021)
Article
Engineering, Mechanical
Pinghe Ni, Qiang Li, Qiang Han, Kun Xu, Xiuli Du
Summary: This study proposes a Bayesian probabilistic model updating approach for substructure identification, which evaluates the uncertainties in identified results by analyzing the responses of large-scale structures. Numerical experiments on a three-span beam structure and experimental studies on an eight-floor steel frame were conducted to verify the accuracy and efficiency of the proposed method, and the results demonstrate its effectiveness.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Computer Science, Artificial Intelligence
Jiahui Wang, Kun Yue, Liang Duan, Zhiwei Qi, Shaojie Qiao
Summary: This paper proposes a Deepwalk-based method for Bayesian network embedding, and approximates probabilistic inferences using the distance among embeddings, providing an efficient approach to multiple probabilistic inferences.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Mathematics, Applied
Oscar Bates, Lluis Guasch, George Strong, Thomas Caradoc Robins, Oscar Calderon-Agudo, Carlos Cueto, Javier Cudeiro, Mengxing Tang
Summary: Bayesian methods are popular for inverse problems research. Stochastic Variational Inference (SVI) uses gradient-based methods to solve Bayes' equation and is suitable for time-limited or computationally expensive applications. SVI can find a cost-free estimate of the pixel-wise variance of the sound-speed distribution, which can be used for uncertainty estimation in pixel-wise sound-speed reconstruction.
Article
Computer Science, Interdisciplinary Applications
Ragheb Raad, Dhruv Patel, Chiao-Chih Hsu, Vijay Kothapalli, Deep Ray, Bino Varghese, Darryl Hwang, Inderbir Gill, Vinay Duddalwar, Assad A. Oberai
Summary: In this manuscript, a novel probabilistic deep-learning algorithm is proposed to impute missing images in a sequence of medical images. The method trains a generative adversarial network (GAN) to learn the underlying probabilistic relation between the images, and uses Bayesian inference to infer the probability distribution of the missing image. A unique style loss for contrast-enhanced computed tomography (CECT) imaging is also proposed to improve the texture of the generated images.
ENGINEERING WITH COMPUTERS
(2022)
Article
Computer Science, Artificial Intelligence
Giso H. Dal, Alfons W. Laarman, Arjen Hommersom, Peter J. F. Lucas
Summary: Bayesian networks are a popular choice for reasoning under uncertainty, but the computational complexity of inference can be a challenge for real-world applications. Weighted Model Counting (WMC) methods aim to reduce the cost of inference by exploiting patterns in the probabilities associated with BN nodes, but require a computationally intensive compilation step. The proposed Compositional Weighted Model Counting (CWMC) framework addresses this issue by partitioning BNs into subproblems, allowing for more efficient application of WMC innovations.
INTERNATIONAL JOURNAL OF APPROXIMATE REASONING
(2021)
Article
Environmental Sciences
Eun-Hee Koh, Eunhee Lee, Kang-Kun Lee, Duk-Cheol Moon
Summary: The combined use of a Bayesian mixing model, numerical model (random walk particle tracking), and environmental tracers was applied to study the groundwater recharge sources, flow path, and residence time in the mountainous area of Jeju Island, South Korea. The study revealed that precipitation during the wet season was the primary source of groundwater, contributing to approximately 64% of the total recharge. Different elevations showed mixed contributions to the recharge sources, and the flow path differed between highland and lowland wells. The study highlights the importance of using integrated analysis techniques to enhance the reliability of recharge area estimation and improve the understanding of complex hydrogeological systems in mountainous areas.
SCIENCE OF THE TOTAL ENVIRONMENT
(2022)