Article
Mathematics, Interdisciplinary Applications
Zachary F. Fisher, Younghoon Kim, Barbara L. Fredrickson, Vladas Pipiras
Summary: Intensive longitudinal data (ILD) is a common data type in social and behavioral sciences, but its potential for forecasting individual-level dynamic processes has not been fully explored. We propose a novel methodological approach, the multi-VAR framework, which can estimate models for multiple individuals simultaneously and adaptively adjust to individual heterogeneity. We introduce a new proximal gradient descent algorithm to solve the multi-VAR problem and prove the consistency of the recovered transition matrices. We evaluate the forecasting performance of our method compared to benchmark methods using a case study of the daily emotional experiences of 16 individuals over 11 weeks.
Article
Engineering, Environmental
Mengya Liu, Qi Li, Fukang Zhu
Summary: This paper proposes a novel categorical time series model to study urban air quality, and analyzes data from three major cities in China. The results indicate that Beijing has the worst air quality but it is gradually improving, and the proposed model shows satisfactory performance through simulation studies using an adaptive Bayesian Markov chain Monte Carlo sampling scheme.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2022)
Article
Automation & Control Systems
Hang Geng, Mulugeta A. Haile, Huazhen Fang
Summary: This article introduces a new method (SSUE) that can simultaneously estimate the internal state and parameter uncertainty of a system to address the challenge of parameter variability in practical dynamic systems. By developing a Bayesian framework and numerical methods, the estimation of parameter uncertainty and the update of the state vector are achieved, while observability analysis is conducted to assess consistency.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2021)
Article
Public, Environmental & Occupational Health
Chuanyun Fu, Tarek Sayed
Summary: Using traffic conflict-based extreme value theory (EVT) models to quantify real-time crash-risk of road facilities is a promising direction for developing proactive traffic safety management strategies. This study proposes a dynamic Bayesian hierarchical peak over threshold modeling approach to estimate real-time crash-risk based on traffic conflicts. The results show that dynamic models considerably outperform static models in terms of statistical fit and predictive performance.
ANALYTIC METHODS IN ACCIDENT RESEARCH
(2023)
Article
Mathematics, Interdisciplinary Applications
Eunsook Kim, Chunhua Cao, Siyu Liu, Yan Wang, Robert Dedrick
Summary: Longitudinal measurement invariance (LMI) is critical for studying change over time with intensive longitudinal data (ILD). In this study, we propose cross-classified factor analysis (CCFA) and alignment optimization (AO) as methods to detect non-invariant item parameters and non-invariant time points, respectively, in LMI testing with ILD. Our results from a Monte Carlo simulation study show that CCFA is an excellent tool for ILD LMI testing, even when autoregression (AR) is misspecified, and can identify a source of non-invariance using a covariate. AO can supplement CCFA in finding non-invariant time points, but it requires a large number of persons.
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Amirreza Farnoosh, Sarah Ostadabbas
Summary: In this paper, we propose a Bayesian switching dynamical model for segmentation of 3D pose data over time that uncovers interpretable patterns in the data and is generative. The experiments demonstrate that our model outperforms state-of-the-art methods on biological motion datasets.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2022)
Article
Neurosciences
Suprateek Kundu, Alec Reinhardt, Serena Song, Joo Han, M. Lawson Meadows, Bruce Crosson, Venkatagiri Krishnamurthy
Summary: This study proposes a novel Bayesian tensor response regression approach for longitudinal neuroimaging data. The method utilizes low-rank decomposition and joint credible regions for feature selection and more accurate inference. The advantages of the proposed approach over traditional voxel-wise regression are highlighted through extensive simulation studies and application to real longitudinal aphasia data.
HUMAN BRAIN MAPPING
(2023)
Article
Astronomy & Astrophysics
Simone Mozzon, Gregory Ashton, Laura K. Nuttall, Andrew R. Williamson
Summary: Gravitational-wave observations of binary neutron star mergers can provide an independent measurement of the Hubble constant, but systematic uncertainties in gravitational-wave observations need to be thoroughly understood. Nonstationary noise in detector data can affect estimations, but it is not expected to be a limiting factor in resolving the tension on H-0 using standard sirens.
Article
Mathematical & Computational Biology
Yuyang Xiao, Juan Shen, Xiufen Zou
Summary: This study presents a mathematical model and dynamical analysis to quantify the hormesis of anti-tumor drugs and determine the critical threshold of antibody dose. The research reveals that low dose antibody promotes tumor growth while high dose antibody inhibits tumor growth. These findings provide suggestions for the appropriate drug dosage in the clinical treatment of cancer.
MATHEMATICAL BIOSCIENCES AND ENGINEERING
(2022)
Article
Substance Abuse
Samuel W. W. Stull, Ashley N. N. Linden-Carmichael, Christy K. K. Scott, Michael L. L. Dennis, Stephanie T. T. Lanza
Summary: Time-varying effect modeling (TVEM) is a statistical technique for modeling dynamic patterns of change, particularly useful when applied to intensive longitudinal data (ILD) to study biobehavioral health processes. TVEM allows flexible modeling of outcomes over continuous time, as well as associations between variables and moderation effects. In the context of addiction research, TVEM coupled with ILD is ideal for studying addiction-related processes.
Article
Clinical Neurology
Sheng Luo, Haotian Zou, Glenn T. Stebbins, Michael A. Schwarzschild, Eric A. Macklin, James Chan, David Oakes, Tanya Simuni, Christopher G. Goetz
Summary: This study applied longitudinal item response theory (IRT) models to analyze the motor symptoms in patients with Parkinson's disease and found distinct progression patterns in tremor and nontremor domains over time and in response to treatment. Levodopa treatment showed a greater slowing effect on tremor severity compared to inosine treatment. Furthermore, longitudinal IRT analysis proved to be a valuable statistical method for monitoring changes in different but related domains.
MOVEMENT DISORDERS
(2022)
Article
Environmental Sciences
Xiaoling Huang, Peng Tian
Summary: The Chinese government has made great efforts and implemented strict regulation policies to reduce air pollutants. Recent studies suggest that these environmental regulations may also contribute to reducing carbon emissions. This paper uses the spatial Durbin model and panel threshold model to investigate the effects of different environmental regulations on carbon emissions reduction based on provincial panel data in China. The findings show that China's net carbon emissions display spatial agglomeration characteristics and that formal and informal environmental regulations have inverted U-shaped impacts on net carbon emissions. The study also reveals that the carbon-reducing effect of environmental regulation becomes more prominent with the improvement of regional technological innovation levels.
JOURNAL OF ENVIRONMENTAL MANAGEMENT
(2023)
Article
Energy & Fuels
Zhenling Chen, Xiaoyan Niu, Xiaofang Gao, Huihui Chen
Summary: This study examines the mechanism underlying the relationship between environmental regulations and green innovation using panel data from 30 Chinese provinces. The results show that market-incentive and public participation-based regulations have positive effects on green innovation, while command-and-control regulations do not. The effects of pollutants on green innovation also vary depending on threshold values. In terms of regions, the eastern region has the greatest impact on promoting green innovation.
FRONTIERS IN ENERGY RESEARCH
(2022)
Article
Multidisciplinary Sciences
Sadegh Arefnezhad, James Hamet, Arno Eichberger, Matthias Fruehwirth, Anja Ischebeck, Ioana Victoria Koglbauer, Maximilian Moser, Ali Yousefi
Summary: This paper presents a novel dynamical modeling solution to estimate the instantaneous level of driver drowsiness using EEG signals. The proposed method provides robust and reliable results in real-time estimation of drowsiness, utilizing biomarkers such as Theta and Delta powers.
SCIENTIFIC REPORTS
(2022)
Article
Environmental Sciences
Weixue Lu, Hecheng Wu, Shuaishuai Geng
Summary: A comprehensive understanding of the heterogeneity and threshold effects of environmental regulation on health expenditure is essential for policy design and decision-making. The study shows significant heterogeneous effects of environmental regulation on health expenditure across different quintiles, as well as region-specific threshold effects in eastern, central, and western China. Stricter environmental regulations are found to reduce health expenditures related to environmental pollution in eastern and western China, but not in central China.
JOURNAL OF ENVIRONMENTAL MANAGEMENT
(2021)