A5052194779- Statistical Methods and Inference
- Bayesian Methods and Mixture Models
- Advanced Statistical Methods and Models
- Statistical Methods and Bayesian Inference
- Metabolomics and Mass Spectrometry Studies
- Computational Drug Discovery Methods
- Advanced Statistical Process Monitoring
- Gaussian Processes and Bayesian Inference
- Fault Detection and Control Systems
- Cancer-related cognitive impairment studies
- Spatial and Panel Data Analysis
- Smart Grid Energy Management
- Energy Load and Power Forecasting
- Statistical Methods in Clinical Trials
- Gene expression and cancer classification
- Spectroscopy and Chemometric Analyses
- Solar Radiation and Photovoltaics
- Dementia and Cognitive Impairment Research
- Voice and Speech Disorders
- Statistical and numerical algorithms
- Insect and Arachnid Ecology and Behavior
- Composting and Vermicomposting Techniques
- Neural Networks and Applications
- Stuttering Research and Treatment
- Sentiment Analysis and Opinion Mining
Michigan State University
2011-2024
Auburn University
2014-2023
University of Science and Technology of China
2008
A polynomial spline estimator is proposed for the mean function of dense functional data together with a simultaneous confidence band which asymptotically correct. In addition, and its accompanying enjoy oracle efficiency in sense that they are same as if all random trajectories observed entirely without errors. The also extended to difference functions two populations data. Simulation experiments provide strong evidence corroborates asymptotic theory while computing efficient. procedure...
Energy management is indispensable in the smart grid, which integrates more renewable energy resources, such as solar and wind. Because of intermittent power generation from these precise forecasting has become crucial to achieve efficient management. In this paper, we propose a novel adaptive learning hybrid model (ALHM) for intensity based on meteorological data. We first present time-varying multiple linear (TMLM) capture dynamic property then construct simultaneous confidence bands...
Core Ideas Meta‐analysis showed that poultry litter’s influence on crop productivity is comparable to of inorganic fertilizer. Poultry effectiveness yield influenced by soil properties, tillage, application practice, and species. More positive effects were found in acidic compared with neutral or alkaline, loam sand clay, under conservation tillage conventional, subsurface banded litter broadcast incorporation through tillage. The full benefits using was achieved from long‐term studies,...
People worldwide use language in subtle and complex ways to express emotions. While emotion recognition -- an umbrella term for several NLP tasks significantly impacts different applications other fields, most work the area is focused on high-resource languages. Therefore, this has led major disparities research proposed solutions, especially low-resource languages that suffer from lack of high-quality datasets. In paper, we present BRIGHTER-- a collection multilabeled emotion-annotated...
In this work, we propose a deep neural networks‐based method to perform non‐parametric regression for functional data. The proposed estimators are based on sparsely connected networks with rectifier linear unit (ReLU) activation function. We provide the convergence rate of estimator in terms empirical norm. Through Monte Carlo simulation studies, examine finite sample performance method. Finally, is applied analyse positron emission tomography images patients Alzheimer's disease obtained...
We consider nonparametric estimation of the covariance function for dense functional data using computationally efficient tensor product B-splines.We develop both local and global asymptotic distributions proposed estimator, show that our estimator is as an "oracle" where true mean known.Simultaneous confidence envelopes are developed based on theory to quantify variability in make inferences covariance.Monte Carlo simulation experiments provide strong evidence corroborates theory.Examples...
We consider nonparametric estimation of the covariance function for dense functional data using computationally efficient tensor product B-splines.We develop both local and global asymptotic distributions proposed estimator, show that our estimator is as an "oracle" where true mean known.Simultaneous confidence envelopes are developed based on theory to quantify variability in make inferences covariance.Monte Carlo simulation experiments provide strong evidence corroborates theory.Two real...
Abstract We propose a new approach, called as functional deep neural network (FDNN), for classifying multidimensional data. Specifically, is trained based on the principal components of training data which shall be used to predict class label future function. Unlike popular discriminant analysis approaches only work one‐dimensional data, proposed FDNN approach applies general non‐Gaussian Moreover, when log density ratio possesses locally connected modular structure, we show that achieves...
Smart Grid is an important component of City, where more renewable power generation and better energy management required. Forecast on generation, from sources such as solar wind, crucial for management. However, the current forecast methods lack a comprehensive understanding natural processes, are thus limited in precise prediction. In this paper, we introduce simultaneous inference to analyze weather data predictions. We first local linear model nonlinear time series, present construction...
In this work, consistent estimators and simultaneous confidence bands for the derivatives of mean functions are proposed when curves repeatedly recorded each subject. The within-curve correlation trajectories has been considered while novel still enjoys semiparametric efficiency. methods lead to a straightforward extension two-sample case in which we compare from two populations. We demonstrate simulations that superior existing approaches ignore dependence. applied investigate mortality...
Translating exercise-science methodology for determination of muscle bioenergetics, we hypothesized that the temporal voice-use patterns classroom and music teachers would indicate a reliance on immediate energy system laryngeal skeletal-muscle metabolism. It was music-teacher group produce longer voiced segments than teachers.Using between- within-group multivariate analysis-of-variance design (5 teachers; 7 teachers), analyzed fundamental-frequency data-collected via an ambulatory...
Neuroimaging studies of cognitive and brain aging often yield massive datasets that create many analytic statistical challenges. In this paper, we discuss address several limitations in the existing work. 1) Linear models are used to model age effects on neuroimaging markers, which inadequate capturing potential nonlinear trends. 2) Marginal correlations network analysis, not efficient characterizing a complex network. 3) Due challenge high- dimensionality, only small subset regional markers...
The intrinsically infinite-dimensional features of the functional observations over multidimensional domains render standard classification methods effectively inapplicable. To address this problem, we introduce a novel multiclass deep neural network (mfDNN) classifier as an innovative data mining and tool. architecture incorporates sparse with Rectified Linear Unit (ReLU) activation function, minimizing cross-entropy loss in framework. This design enables utilization modern computational...
In this paper, we study the estimation for generalized partially linear single-index models, where systematic component in model has a flexible semi-parametric form with general link function. We propose an efficient and practical approach to estimate function, coefficients as well of model. The procedure is developed by applying quasi-likelihood polynomial spline smoothing. derive large sample properties estimators show convergence rate each Asymptotic normality semiparametric efficiency...
Abstract We propose a score‐type statistic to evaluate heterogeneity in zero‐inflated models for count data stratified population, where is defined as instances which the zero counts are generated from two sources. Evaluating this class of has attracted considerable attention literature, but existing testing procedures have primarily relied on constancy assumption under alternative hypothesis. In paper, we extend literature by describing test homogeneity against general alternatives that do...
A central topic in functional data analysis is how to design an optimal decision rule, based on training samples, classify a function.We exploit the classification problem when functions are Gaussian processes.Sharp convergence rates for minimax excess misclassification risk derived both settings that fully observed and discretely observed.We explore two easily implementable classifiers discriminant deep neural network, respectively, which proven achieve optimality settings.Our network...
In this paper, we present a class of functional linear regression models with varying coefficients response on one or multiple predictors and scalar predictors. particular, the approach can accommodate densely sparsely sampled responses as well It also allows for combination continuous categorical covariates. Tensor product B‐spline basis is proposed estimation bivariate coefficient functions. We show that our estimators hold asymptotic consistency normality. Several numerical examples...