Article
Mathematics, Interdisciplinary Applications
Baolei Wei
Summary: The study proposes an integral SINDy (ISINDy) method to identify the model structure and parameters of nonlinear ordinary differential equations (ODEs) from noisy time-series observations. The approach combines penalized spline smoothing and sequential threshold least squares to achieve sparse pseudo linear regression and extract the fewest active terms. The method shows accuracy and robustness to noise in various simulations.
CHAOS SOLITONS & FRACTALS
(2022)
Article
Mathematical & Computational Biology
Minggen Lu, Yan Liu, Chin-Shang Li, Jianguo Sun
Summary: Efficient estimation of flexible transformation models with interval-censored data is studied in this paper. To reduce the dimension of semiparametric models, the unknown monotone function is approximated using a monotone B-spline. A penalization technique is used to computationally efficiently estimate all parameters. An easy-to-implement nested iterative expectation-maximization (EM) algorithm is developed for estimation and a simple variance-covariance estimation approach is proposed for large-sample inference of the regression parameters. Theoretical results show that the estimator achieves the optimal convergence rate for the unknown monotone increasing function, and the estimators of the regression parameters are asymptotically normal and efficient under appropriate selection of the smoothing parameter order and the spline space knots. Extensive numerical experiments and implementation in the R package PenIC assess the proposed penalized procedure. The methodology is further illustrated through a signal transduction study.
STATISTICS IN MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Yimiao Gao, Yuehan Yang
Summary: This paper proposes a method called JETS that utilizes auxiliary models from different groups to estimate the target model. By constructing a penalized framework that combines penalties for the target model and the differences between auxiliary models and the target model, JETS overcomes the challenge of limited samples in high-dimensional studies and obtains stable and accurate estimates, regardless of noisy information in the auxiliary samples.
PATTERN RECOGNITION
(2023)
Article
Statistics & Probability
Yan Liu, Minggen Lu, Christopher S. McMahan
Summary: A partially linear additive transformation model is proposed for analyzing current status data, using constrained B-splines to model monotone transformation functions and nonparametric covariate effects. A penalization technique is utilized for more efficient estimators, and an easy to implement hybrid algorithm is developed for model fitting. The proposed estimators are shown to have excellent finite-sample performance and convergence rates in theoretical analysis.
ELECTRONIC JOURNAL OF STATISTICS
(2021)
Review
Chemistry, Analytical
Francisco Raposo, Damia Barcelo
Summary: This critical review paper discusses the main analytical calibration models and their practical use guidelines. It proposes a three-step simple calibration diagnosis method based on a combination of graphical plots, statistical significance tests, and numerical parameters. Experimental conditions and calibration procedure design are crucial for the appropriate selection of models.
TRAC-TRENDS IN ANALYTICAL CHEMISTRY
(2021)
Article
Computer Science, Theory & Methods
Xinmin Li, Haozhe Liang, Wolfgang Haerdle, Hua Liang
Summary: We propose test statistics based on the penalized spline to decide between generalized linear models and generalized partially linear models. The numerical performance of the proposed statistics is comparable to that of their kernel-based competitors, which have been shown to be asymptotically normal in the literature (Hardle et al. in J Am Stat Assoc 93:1461-1474, 1998). We also numerically explore the possibility of using the proposed statistics for goodness of fit checking for GLM. The proposed procedures are illustrated to analyze two datasets.
STATISTICS AND COMPUTING
(2023)
Article
Computer Science, Interdisciplinary Applications
Vimukthini Pinto, Roshini Sooriyarachchi
Summary: Multilevel modelling is a novel approach for analyzing data with hierarchical structures. This study compares estimation methods for a goodness-of-fit test developed for binary response multilevel models, based on mathematical background, extensive simulations, and application to real-life data.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2021)
Article
Statistics & Probability
Takuma Yoshida
Summary: This paper focuses on establishing the asymptotic distribution of the quantile estimator obtained by the penalized spline method. Simulation data and real data are used for empirical analysis to validate our results.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2023)
Article
Biology
X. Nie, S. Wager
Summary: The article introduces a general two-step algorithm for estimating heterogeneous treatment effects, which is flexible and easy to use, and has several advantages over existing methods, showing promising performance in simulation setups.
Article
Biology
Timo Dimitriadis, Lutz Dumbgen, Alexander Henzi, Marius Puke, Johanna Ziegel
Summary: Probability predictions from binary regressions or machine learning methods should be calibrated. We propose an honest calibration assessment based on novel confidence bands for the calibration curve, which are valid subject to only the natural assumption of isotonicity. In an application to modeling predictions of an infant having low birth weight, the bounds give informative insights into model calibration.
Article
Engineering, Environmental
M. Ross Kunz, Adam Yonge, Zongtang Fang, Rakesh Batchu, Andrew J. Medford, Denis Constales, Gregory Yablonsky, Rebecca Fushimi
Summary: Understanding the elementary steps and kinetics in each reaction is crucial for decision-making about catalytic materials. This work introduces a methodology combining transient rate/concentration dependencies and machine learning to measure active sites, rate constants, and investigate reaction mechanisms under complex steps. The approach can provide accurate estimates of micro-kinetic coefficients and reveal how materials control reaction mechanisms through experimental data.
CHEMICAL ENGINEERING JOURNAL
(2021)
Article
Statistics & Probability
Yoonsuh Jung, Steven N. MacEachern, Hang Kim
Summary: This study proposes a modified check loss function with quadratic adjustment to guard against overfitting, which has been empirically proven through various simulation settings of linear and nonlinear regressions.
JOURNAL OF APPLIED STATISTICS
(2021)
Article
Statistics & Probability
Ana Perez-Gonzalez, Tomas R. Cotos-Yanez, Wenceslao Gonzalez-Manteiga, Rosa M. Crujeiras-Casais
Summary: This paper introduces and analyzes goodness-of-fit tests for quantile regression models in the presence of missing observations in the response variable, based on construction of empirical processes and considering three different approaches. The performance of different test statistics is extensively studied through simulation, with an application to real data included.
STATISTICAL PAPERS
(2021)
Article
Engineering, Multidisciplinary
Yaoyao He, Huiling Fan, Xiaohui Lei, Jinhong Wan
Summary: This paper presents a B-spline quantile regression probability density prediction method for accurate runoff forecasting and quantifying uncertainty. By constructing a probability density curve and evaluating point and interval predictions, the method performs well in runoff prediction.
APPLIED MATHEMATICAL MODELLING
(2021)
Article
Computer Science, Information Systems
Ruiting Hao, Qiwei Han, Lu Li, Xiaorong Yang
Summary: This study proposes an iterative estimation method based on the data augmentation algorithm for censored MCQRNN model. Simulation studies and real data application demonstrate that our proposed method outperforms existing censored methods in terms of quantile loss and C-index, and yields prediction results very close to those obtained from full uncensored data. By avoiding the quantile crossing problem, the proposed method is more flexible and can handle different types of censoring, overcoming the limitation of existing censored methods that only work for single censoring type with unreliable predictions caused by the crossing phenomenon.
INFORMATION SCIENCES
(2023)
Article
Statistics & Probability
Minggen Lu, Tao Lu, Chin-Shang Li
JOURNAL OF STATISTICAL PLANNING AND INFERENCE
(2018)
Article
Psychology, Developmental
Taylor Lensch, Kristen Clements-Nolle, Roy F. Oman, William P. Evans, Minggen Lu, Wei Yang
Summary: The study shows that family communication and school connectedness can offer direct protection against suicidal behaviors in the presence of adverse childhood experiences, as well as buffer the association between ACEs and suicidal behaviors on the multiplicative scale. These findings support the development of interventions to enhance family communication and school connectedness to reduce suicidal behaviors, and screening for trauma and suicidal behaviors is recommended.
JOURNAL OF ADOLESCENT HEALTH
(2021)
Article
Mathematical & Computational Biology
Minggen Lu, Yan Liu, Chin-Shang Li, Jianguo Sun
Summary: Efficient estimation of flexible transformation models with interval-censored data is studied in this paper. To reduce the dimension of semiparametric models, the unknown monotone function is approximated using a monotone B-spline. A penalization technique is used to computationally efficiently estimate all parameters. An easy-to-implement nested iterative expectation-maximization (EM) algorithm is developed for estimation and a simple variance-covariance estimation approach is proposed for large-sample inference of the regression parameters. Theoretical results show that the estimator achieves the optimal convergence rate for the unknown monotone increasing function, and the estimators of the regression parameters are asymptotically normal and efficient under appropriate selection of the smoothing parameter order and the spline space knots. Extensive numerical experiments and implementation in the R package PenIC assess the proposed penalized procedure. The methodology is further illustrated through a signal transduction study.
STATISTICS IN MEDICINE
(2022)
Article
Statistics & Probability
Minggen Lu, Chin-Shang Li, Karla D. Wagner
Summary: We developed a practical and computationally efficient penalised estimation approach for partially linear additive models to deal with zero-inflated binary outcome data. The approach utilizes B-splines to approximate unknown nonparametric components and employs a two-stage iterative expectation-maximisation (EM) algorithm to calculate penalised spline estimates. We established the large-sample properties such as uniform convergence and optimal rate of convergence for functional estimators, as well as asymptotic normality and efficiency for regression coefficient estimators.
JOURNAL OF NONPARAMETRIC STATISTICS
(2023)