Article
Engineering, Electrical & Electronic
Long Shi, Ruyuan Lu, Zhuofei Liu, Jiayi Yin, Ye Chen, Jun Wang, Lu Lu
Summary: This article introduces the importance of time-series prediction and the application of kernel adaptive filtering. By proposing a robust KAF algorithm, it effectively addresses the performance deterioration problem in nonlinear regression tasks. The convergence properties and steady-state excess MSE of the algorithm are analyzed to derive the condition for determining the appropriate step size.
IEEE SENSORS JOURNAL
(2023)
Article
Energy & Fuels
Faisal Aladwani, Adel Elsharkawy
Summary: Fluid viscosity is crucial in the petroleum industry. This study developed three supervised machine learning regression models to predict viscosity and found that the newly developed models outperformed existing models, with Gaussian process regression and regression ensembles tree showing the best performance.
PETROLEUM SCIENCE AND TECHNOLOGY
(2023)
Article
Mechanics
Su Tian, Wenbin Yu
Summary: In this work, the Gaussian process is proposed to approximate the overall shape of the initial failure envelope in the space of sectional forces and moments of composite beam cross-sections, using Tsai-Wu failure criterion for strength analysis at the material level. The adaptive sampling technique is used to select training data and the predicted variance scaled by predicted mean is used as a measure of the score to select the most promising point as the new sample for the next training iteration.
COMPOSITE STRUCTURES
(2021)
Article
Computer Science, Artificial Intelligence
Tanner Norton, Grant Stagg, Derek Ward, Cameron K. Peterson
Summary: In this paper, a decentralized sparse Gaussian process regression (DSGPR) model with event-triggered, adaptive inducing points is presented. The paper addresses common issues in decentralized frameworks such as high computation costs, message bandwidth restrictions, and data fusion integrity. An improvement to real-time sparse Gaussian process regression models is introduced by using radial clustering for adaptive inducing point selection. The DSGPR framework is evaluated on simulated random vector fields, demonstrating its effectiveness in estimating vector fields using multiple autonomous agents.
JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Yiqi Liu, Daoping Huang, Bin Liu, Qiang Feng, Baoping Cai
Summary: This paper introduces a novel ensemble learning algorithm that combines global and local GPR models to accurately predict quality-related variables in industrial processes. An adaptive ranking strategy and variable selection method further enhance the accuracy of predictions.
APPLIED SOFT COMPUTING
(2021)
Article
Mathematics
Shaohao Xie, Shaohua Zhuang, Yusong Du
Summary: This paper revisits the centered discrete Gaussian distributions Bernoulli sampling proposed by Ducas et al. in 2013, presenting an improved algorithm and a noncentered version for varying centers over the integers. Experimental results show a significant improvement in sampling efficiency compared to other rejection algorithms.
Article
Engineering, Industrial
Atin Roy, Subrata Chakraborty
Summary: In this study, a three-stage adaptive support vector regression (SVR) model is built to alleviate the scarcity of samples in the reliability evaluation of structures with implicit limit state functions (LSFs). The model employs sequential and importance sampling techniques to ensure a sufficient number of simulation points near the failure plane for accurate estimation of reliability.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2022)
Article
Geosciences, Multidisciplinary
Maryam Sadi, Abbas Shahrabadi
Summary: In this study, experimental measurements and modeling investigations were conducted to predict crude oil viscosity under various conditions. Three advanced intelligent models were developed to estimate saturated and under-saturated oil viscosity using input parameters such as crude oil API, solution gas oil ratio, bubble point pressure, molecular weight, specific gravity of C12+ fraction, mole percent of C?11components, temperature, and pressure. The results showed that the Gaussian process regression model had the best performance in viscosity prediction, with average absolute relative errors of 0.18% and 0.07% for saturated and under-saturated oil, respectively. The findings of the Leverage technique and sensitivity analysis further supported the reliability and importance of the study.
NATURAL RESOURCES RESEARCH
(2023)
Article
Mechanics
Zhibao Cheng, Min Li, Gaofeng Jia, Zhifei Shi
Summary: This paper proposes an adaptive Gaussian process (AGP) model to efficiently predict the complex dispersion relations for periodic structures. It first predicts the coefficients of the dispersion equation at selected frequencies, and then analytically solves the dispersion equation to establish the complex dispersion relation. PCA is used to reduce the dimension of these coefficients, and an adaptive procedure is integrated to improve the accuracy of the GP model.
EUROPEAN JOURNAL OF MECHANICS A-SOLIDS
(2022)
Article
Engineering, Multidisciplinary
Zhaoyi Xu, Yanjie Guo, Joseph H. Saleh
Summary: This article proposes a hybrid Bayesian BFGS algorithm (HB2O) to address the efficiency problem, and develops an adaptive expected improvement (AEI) acquisition function to realize a self-adaptive sampling strategy. The computational experiments demonstrate that the HB2O can efficiently converge on functions' optima with limited simulation samples and outperform other optimizers for various test functions.
ENGINEERING OPTIMIZATION
(2021)
Article
Energy & Fuels
Jiazhi Liu, Xintian Liu
Summary: This study proposes a data-driven algorithm to predict the functional capacity of Li-ion batteries, using the health factor (HF) and weighted particle swarm optimization Gaussian process regression (WPSO-GPR) models to predict the health status of batteries under different temperature conditions. The HF is adaptively extracted from the voltage and time profiles during discharge stages, while a GPR model with rational quadratic coefficients as the kernel function is used for capacity regeneration. An adaptive weight particle swarm algorithm is employed to optimize the GPR model. The proposed framework is validated using two evaluation metrics: root mean square error (RMSE) and mean absolute percentage error (MAPE), showing high prediction accuracy and broad applicability.
JOURNAL OF ENERGY STORAGE
(2023)
Article
Green & Sustainable Science & Technology
Shuaihua Shen, Yanxuan Du, Zhengjie Xu, Xiaoqiang Qin, Jian Chen
Summary: In this paper, an SVR adaptive optimization rolling composite model with the STOA algorithm is proposed for temperature prediction. An adaptive Gauss-Cauchy mutation operator is introduced to increase population diversity and search space, and the improved algorithm optimizes the key parameters of the SVR model. Rolling prediction is integrated into the model and real-time update principles are used to continuously update the prediction, improving accuracy. The model is validated using global mean temperature data and compared with other models, showing better prediction performance.
Article
Agricultural Engineering
Sahand Assadzadeh, Cassandra K. Walker, Linda S. McDonald, Joe F. Panozzo
Summary: This study constructed predictive models of flour milling yield (MY)% using image and spectral data. The models, including multiple linear regression, support vector regression, and Gaussian process regression, were built based on extracted features from the data. The results showed that the combination of all features as input variables for a Gaussian process regression model achieved the best accuracy.
BIOSYSTEMS ENGINEERING
(2022)
Article
Green & Sustainable Science & Technology
Haoshu Cai, Xiaodong Jia, Jianshe Feng, Qibo Yang, Wenzhe Li, Fei Li, Jay Lee
Summary: This paper introduces a unified filtering framework for wind speed prediction that integrates short-term prediction model, Numerical Weather Prediction, and a smoothing term. The proposed framework shows improved prediction accuracy in short-term horizon and a 46% enhancement in RMSE accuracy for medium/long-term horizon compared to benchmarks using offshore wind farm data.
Article
Food Science & Technology
Fuxiang Wang, Chunguang Wang
Summary: In this study, visible-near-infrared (VIS-NIR) hyperspectral imaging combined with a data fusion strategy was used for the nondestructive assessment of starch content in intact potatoes. The results showed that appropriate mid-level data fusion can effectively improve the prediction accuracy for starch content and aid in sorting potato starch content in the production line.
Article
Physics, Applied
Yuki K. Wakabayashi, Kohei Okamoto, Yoshisuke Ban, Shoichi Sato, Masaaki Tanaka, Shinobu Ohya
APPLIED PHYSICS EXPRESS
(2016)
Article
Nanoscience & Nanotechnology
Kosuke Takiguchi, Yuki K. Wakabayashi, Kohei Okamoto, Masaaki Tanaka, Shinobu Ohya
Article
Physics, Applied
Ryota Suzuki, Yuki K. Wakabayashi, Kohei Okamoto, Masaaki Tanaka, Shinobu Ohya
APPLIED PHYSICS LETTERS
(2018)
Article
Physics, Applied
Tae-Eon Bae, Yuki Wakabayashi, Ryosho Nakane, Mitsuru Takenaka, Shinichi Takagi
JAPANESE JOURNAL OF APPLIED PHYSICS
(2018)
Article
Physics, Applied
Yoshisuke Ban, Yuki K. Wakabayashi, Ryosho Nakane, Masaaki Tanaka
JOURNAL OF APPLIED PHYSICS
(2018)
Article
Materials Science, Multidisciplinary
Yuki K. Wakabayashi, Yosuke Nonaka, Yukiharu Takeda, Shoya Sakamoto, Keisuke Ikeda, Zhendong Chi, Goro Shibata, Arata Tanaka, Yuji Saitoh, Hiroshi Yamagami, Masaaki Tanaka, Atsushi Fujimori, Ryosho Nakane
PHYSICAL REVIEW MATERIALS
(2018)
Article
Multidisciplinary Sciences
Yuki K. Wakabayashi, Yoshiharu Krockenberger, Naoto Tsujimoto, Tommy Boykin, Shinji Tsuneyuki, Yoshitaka Taniyasu, Hideki Yamamoto
NATURE COMMUNICATIONS
(2019)
Article
Nanoscience & Nanotechnology
Yuki K. Wakabayashi, Takuma Otsuka, Yoshiharu Krockenberger, Hiroshi Sawada, Yoshitaka Taniyasu, Hideki Yamamoto
Review
Physics, Applied
Yuki K. Wakabayashi, Yoshiharu Krockenberger, Takuma Otsuka, Hiroshi Sawada, Yoshitaka Taniyasu, Hideki Yamamoto
Summary: This review summarizes the methods of growing ultra-high-quality SrRuO3 films and their novel physics, as well as discusses the progress in crystal structure analyses and the electrical and magnetic properties of SrRuO3 over the last decade.
JAPANESE JOURNAL OF APPLIED PHYSICS
(2023)
Article
Materials Science, Multidisciplinary
Shoya Sakamoto, Yosuke Nonaka, Keisuke Ikeda, Zhendong Chi, Yuxuan Wan, Masahiro Suzuki, Atsushi Fujimori, Le Duc Anh, Pham Nam Hai, Yukiharu Takeda, Yuji Saitoh, Masaki Kobayashi, Masaaki Tanaka, Yuki K. Wakabayashi, Hiroshi Yamagami
Article
Materials Science, Multidisciplinary
Shinobu Ohya, Akiyori Yamamoto, Tomonari Yamaguchi, Ryo Ishikawa, Ryota Akiyama, Le Duc Anh, Shobhit Goel, Yuki K. Wakabayashi, Shinji Kuroda, Masaaki Tanaka
Article
Materials Science, Multidisciplinary
Yuki K. Wakabayashi, Yosuke Nonaka, Yukiharu Takeda, Shoya Sakamoto, Keisuke Ikeda, Zhendong Chi, Goro Shibata, Arata Tanaka, Yuji Saitoh, Hiroshi Yamagami, Masaaki Tanaka, Atsushi Fujimori, Ryosho Nakane
Article
Materials Science, Multidisciplinary
S. Sakamoto, Y. K. Wakabayashi, Y. Takeda, S. -I. Fujimori, H. Suzuki, Y. Ban, H. Yamagami, M. Tanaka, S. Ohya, A. Fujimori
Article
Materials Science, Multidisciplinary
Yuki K. Wakabayashi, Ryota Akiyama, Yukiharu Takeda, Masafumi Horio, Goro Shibata, Shoya Sakamoto, Yoshisuke Ban, Yuji Saitoh, Hiroshi Yamagami, Atsushi Fujimori, Masaaki Tanaka, Shinobu Ohya