Article
Multidisciplinary Sciences
Xin Gao, Frank Konietschke, Qiong Li
Summary: Simultaneous confidence intervals are proposed based on the percentile bootstrap approach, with demonstrated admissibility and correct coverage probability for both normal and non-normal distributions. Efron's nonparametric delta method is extended for constructing nonparametric simultaneous confidence intervals for LASSO regression estimates.
Article
Engineering, Multidisciplinary
Samuel Rosa, Andrea Kocianova
Summary: This paper addresses the issue of accuracy in estimating critical gaps at unsignalized intersections, providing new methods for measuring the accuracy of such estimates and offering guidance on the number of observations required for reliable results.
ALEXANDRIA ENGINEERING JOURNAL
(2023)
Article
Biochemistry & Molecular Biology
David R. Bickel
Summary: Confidence intervals of divergence times and branch lengths do not reflect uncertainty about their clades or model assumptions. Uncertainty about clades can be adjusted by multiplying confidence level with bootstrap proportion, while uncertainty about the model can be propagated by reporting the union of confidence intervals from plausible models. The proposed methods of uncertainty quantification may be used together.
MOLECULAR PHYLOGENETICS AND EVOLUTION
(2022)
Article
Biochemistry & Molecular Biology
David R. Bickel
Summary: This paper demonstrates how to propagate the uncertainty of assumptions by combining multiple models in order to construct more accurate phylogenetic trees from molecular sequence data. By considering all assumptions and their prior distributions, the combined models incorporate much more uncertainty than Bayesian model averages, therefore avoiding overly confident conclusions. The proposed model combination method is validated using nucleotide sequence data.
MOLECULAR PHYLOGENETICS AND EVOLUTION
(2023)
Article
Multidisciplinary Sciences
Wisunee Puggard, Sa-Aat Niwitpong, Suparat Niwitpong
Summary: This study introduces four methods for constructing confidence intervals for the coefficient of variation (CV) and the difference between CVs of Birnbaum-Saunders (BS) distributions. A Monte Carlo simulation shows that the highest posterior density (HPD) interval performs best overall. The proposed methods were validated using PM 2.5 concentration data for Chiang Mai, Thailand in March and April 2019, with results consistent with the simulation findings.
Article
Multidisciplinary Sciences
Wasana Chankham, Sa-Aat Niwitpong, Suparat Niwitpong
Summary: This study proposed confidence intervals for the coefficient of variation (CV) of an inverse Gaussian (IG) distribution to calculate PM2.5 concentrations. The performance of these intervals was evaluated through simulation experiments and it was found that the fiducial confidence interval (FCI) performed the best, especially for small sample sizes.
Article
Computer Science, Interdisciplinary Applications
Zhengguo Xu, Matilde Merino-Sanjuan, Victor Mangas-Sanjuan, Alfredo Garcia-Arieta
Summary: This study investigated the suitability of five estimators and fourteen types of confidence intervals for comparing dissolution profiles through simulation. The best combination to control type I errors was found to be using the estimator calculated based on the mathematical expectation of (f) over cap (2), and the variance-corrected estimator, combined with any of the ten percentile intervals.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Article
Construction & Building Technology
Yuwei Jin, Moses Amoasi Acquah, Mingyu Seo, Sekyung Han
Summary: This study proposes a CNN-GRU hybrid model with parameter-based transfer learning to enhance prediction accuracy for small datasets; utilizing the STE method to estimate data distribution bandwidth and provide load prediction intervals, demonstrating the effectiveness of the method in load forecasting.
ENERGY AND BUILDINGS
(2022)
Article
Engineering, Industrial
Liangliang Zhuang, Ancha Xu, Jihong Pang
Summary: A novel two-stage method called fractional-random-weight bootstrap is proposed to help make interval estimation for both model parameters and future failure numbers in the presence of heavy censoring and batch effects. The method demonstrates superiority over other commonly-used bootstrap methods in simulation studies, especially when heavy censoring is present. Application of the methodology to a real dataset highlights the importance of considering batch effects in interval data to avoid inaccurate predicted number of failures.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2021)
Article
Psychology, Mathematical
Junyeong Yang, Jiwon Kim, Minjung Kim
Summary: This study provides an overview of various methods for detecting dyadic patterns in the actor-partner interdependence model (APIM). By evaluating and comparing four different methods, it was found that the new-variable approach and chi(2) difference test performed better in detecting dyadic patterns. The findings suggest a novel procedure for examining dyadic patterns in APIM.
BEHAVIOR RESEARCH METHODS
(2023)
Article
Engineering, Industrial
Zequan Chen, Guofa Li, Jialong He, Zhaojun Yang, Jili Wang
Summary: This paper introduces a novel adaptive structural reliability analysis strategy, called confidence interval squeezing (CIS) method, based on the Kriging method. The CIS method aims to improve the estimation accuracy of failure probability by squeezing the confidence interval at the highest speed possible, and proposes a new global convergence condition based on the confidence interval of failure probability. In addition, three variants of the CIS method are proposed to handle the difficulty of direct application. Several examples demonstrate that each CIS method can efficiently and accurately handle complex limit state functions and engineering problems with implicit functions.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2022)
Article
Engineering, Mechanical
Xun Wang, Mohamed S. Ghidaoui, Jing Lin
Summary: The study proposes a leak localization method that can accurately predict the location of a leak and provide additional information about uncertainty by introducing confidence intervals and empirical distributions.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Statistics & Probability
T. Tony Cai, Zijian Guo, Rong Ma
Summary: This article presents a unified statistical inference framework for high-dimensional binary generalized linear models (GLMs), considering both unknown and known design distribution settings. A two-step weighted bias-correction method is proposed to construct confidence intervals (CIs) and simultaneous hypothesis tests for individual components of the regression vector. The proposed method is shown to be effective and superior through simulation studies and analysis of real data, providing important insights on the sparsity of the regression vector.
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
(2023)
Article
Ecology
David Fletcher, Tim Jowett
Summary: This paper presents a simple alternative to the delta method for calculating confidence intervals of nonlinear functions of model parameters. The proposed method performs better when the estimates of model parameters have an approximate multivariate normal distribution, and a diagnostic tool is provided to assess the reliability of this assumption. Compared to other methods, it is computationally efficient and provides better coverage properties without the need to refit the model.
METHODS IN ECOLOGY AND EVOLUTION
(2022)
Article
Computer Science, Interdisciplinary Applications
Han Lin Shang
Summary: The study introduces a double bootstrap procedure to reduce coverage error in confidence intervals of descriptive statistics for functional data, outperforming single bootstrap in finite sample performance. Despite longer computational time, the double bootstrap method improves coverage accuracy compared to the single bootstrap.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2021)
