Linear Data Models Research Articles

The sparse estimation of the semiparametric linear transformation model with dependent current status data

In this paper, we study the sparse estimation under the semiparametric linear transformation models for the current status data, also called type I interval-censored data. For the problem, the failure time of interest may be dependent on the censoring time and the association parameter between them is left unspecified. To address this, we employ the copula model to describe the dependence between them and a two-stage estimation procedure to estimate both the association parameter and the regression parameter. In addition, we propose a penalized maximum likelihood estimation procedure based on the broken adaptive ridge regression, and Bernstein polynomials are used to approximate the nonparametric functions involved. The oracle property of the proposed method is established and the numerical studies suggest that the method works well for practical situations. Finally, the method is applied to an Alzheimer's disease study that motivated this investigation.

Slepian-Bangs formulas for parameterized density generator of elliptically symmetric distributions

This paper mainly deals with an extension of the matrix Slepian-Bangs (SB) formula to elliptical symmetric (ES) distributions under the assumption that the arbitrary density generator depends on unknown parameters, aiming to rigorously quantify and understand the impact of this assumption on ES distributed parametric estimation models. This matrix SB formula is derived in a unified way within the framework of real (RES) and circular (C-CES) or noncircular (NC-CES) complex elliptically symmetric distributions, and then compared to the matrix SB formula obtained with fully known or completely unknown density generators. This new matrix SB formula involves a common structure to the existing one with a simple corrective coefficient. Closed-form expressions of this coefficient are given for Student’s t and generalized Gaussian distributions and are each compared according to different knowledge of the density generator. This allows us to conclude that for an arbitrary parameterization, the Cramér-Rao bound (CRB) may be very sensitive to the knowledge of the density generator for super-Gaussian distributions contrary to sub-Gaussian distributions. Finally, we prove that for the parametrization with an unknown scale factor, the CRB for the estimation of the other parameters of the scatter matrix does not depend on the type of knowledge of the density generator. This latter result remains true for the specific noisy linear mixture data model where the parameter of interest is characterized by the range space of the mixing matrix.

A CHEMTAX Study Based on Picoeukaryotic Phytoplankton Pigments and Next-Generation Sequencing Data from the Ulleungdo–Dokdo Marine System of the East Sea (Japan Sea): Improvement of Long-Unresolved Underdetermined Bias

The CHEMTAX program has been widely used to estimate community composition based on major pigment concentrations in seawater. However, because CHEMTAX is an underdetermined optimization algorithm, underdetermined bias has remained an unsolved problem since its development in 1996. The risk of producing biased results increases when analyzing the picophytoplankton community; therefore, this study tested a new method for avoiding biased CHEMTAX results using the picophytoplankton community around the East Sea (Japan Sea). This method involves building a linear model between pigment concentration data and community composition data based on DNA sequencing to predict the pigment range for each operational taxonomic unit, based on the 95% prediction interval. Finally, the range data are transformed into an initial ratio and ratio limits for CHEMTAX analysis. Three combinations of initial ratios and ratio limits were tested to determine whether the modeled initial ratio and ratio limit could prevent underdetermined bias in the CHEMTAX estimates; these combinations were the modeled initial ratio and ratio limit, the modeled initial ratio with a default ratio limit of 500 s, and an initial ratio from previous research with the default ratio limit. The final ratio and composition data for each combination were compared with Bayesian compositional estimator-based final ratio and composition data, which are robust against underdetermined bias. Only CHEMTAX analysis using the modeled initial ratio and ratio limit was unbiased; all other combinations showed significant signs of bias. Therefore, the findings in this study indicate that ratio limits and the initial ratio are equally important in the CHEMTAX analysis of biased datasets. Moreover, we obtained statistically supported initial ratios and ratio limits through linear modeling of pigment concentrations and 16s rDNA composition data.

Autohrf-an R package for generating data-informed event models for general linear modeling of task-based fMRI data.

The analysis of task-related fMRI data at the level of individual participants is commonly based on general linear modeling (GLM), which allows us to estimate the extent to which the BOLD signal can be explained by the task response predictors specified in the event model. The predictors are constructed by convolving the hypothesized time course of neural activity with an assumed hemodynamic response function (HRF). However, our assumptions about the components of brain activity, including their onset and duration, may be incorrect. Their timing may also differ across brain regions or from person to person, leading to inappropriate or suboptimal models, poor fit of the model to actual data, and invalid estimates of brain activity. Here, we present an approach that uses theoretically driven models of task response to define constraints on which the final model is computationally derived using actual fMRI data. Specifically, we developed autohrf-an R package that enables the evaluation and data-driven estimation of event models for GLM analysis. The highlight of the package is the automated parameter search that uses genetic algorithms to find the onset and duration of task predictors that result in the highest fitness of GLM based on the fMRI signal under predefined constraints. We evaluated the usefulness of the autohrf package on two original datasets of task-related fMRI activity, a slow event-related spatial working memory study and a mixed state-item study using the flanker task, and on a simulated slow event-related working memory data. Our results suggest that autohrf can be used to efficiently construct and evaluate better task-related brain activity models to gain a deeper understanding of BOLD task response and improve the validity of model estimates. Our study also highlights the sensitivity of fMRI analysis with GLM to precise event model specification and the need for model evaluation, especially in complex and overlapping event designs.

Robust model‐free adaptive longitudinal flight control for a flapping wing micro air vehicle with wind disturbances

AbstractIn this paper, two robust model‐free adaptive control (RMFAC) methods are proposed for a nonlinear flapping wing micro air vehicle (FWMAV) subject to measurable and unmeasurable wind disturbances. Firstly, a novel disturbance‐related full‐form dynamic linearization technique (DFFDL) is developed to transform the FWMAV with measurable disturbances into a dynamic linearization data model. Then, a DFFDL based RMFAC method is designed based on the obtained data model for the FWMAV to suppress the wind disturbances. Next, the stability analysis shows that the system output tracking error converges into a bounded range. And the above result is extended to the case of unmeasurable disturbances. Finally, the simulation comparison results demonstrate the effectiveness of two proposed RMFAC methods.

Bayesian variable selection and estimation in multivariate skew-normal generalized partial linear mixed models for longitudinal data

This paper proposes a novel generalized partial linear mixed models (GPLMMs) with the random effects distribution distributed as a multivariate skew-normal distribution for discrete longitudinal data over time. Under the Bayesian framework, we extended the Bayesian version of adaptive lasso (least absolute shrinkage and selection operator) for linear regression to the Bayesian adaptive lasso with the least squares approximation (LSA) method for generalized partial linear mixed models. A hybrid algorithm combining the Gibbs sampler and the Metropolis-Hastings algorithm in conjunction with the Bayesian adaptive lasso with LSA is used to simultaneously obtain the Bayesian estimates of unknown parameters, smoothing function and random effects, as well as select important covariates in GPLMMs. Several simulation studies and a real example are presented to illustrate the proposed methodologies.

COVID-19 vaccination acceptance among community members and health workers in Ebonyi state, Nigeria: study protocol for a concurrent-independent mixed method analyses of intention to receive, timeliness of the intention to receive, uptake and hesitancy to COVID-19 vaccination and the determinants

IntroductionThe COVID-19 pandemic has gravely affected the lives and economies of the global population including Nigeria. The attainment of herd immunity through mass COVID-19 vaccination is the foremost control strategy,...

Genetic parameters for dry matter intake, energy balance, residual energy intake, and liability to diseases in German Holstein and Fleckvieh dairy cows.

Selection for feed efficiency (FE) is a hot topic in dairy cow breeding. Dry matter intake (DMI) and residual energy intake (REI) are mostly discussed as new selection traits. Selection for lower DMI or REI seems to increase FE if other traits, such as milk yield or health, are not affected negatively. However, genetic relationships with other traits have not been adequately investigated because of the difficulties in recording sufficient feed intake data for genetic evaluations. The aim of this study was to examine the genetic relationships between FE-related traits and liability to diseases throughout lactation. First, heritabilities for all traits are presented. Subsequently, genetic correlations between DMI, energy-corrected milk yield, energy balance (EB), and REI on the one hand and 3 disease categories (mastitis, claw and leg diseases, and all diseases) on the other throughout lactation in German Holstein (GH) dairy cows are illustrated. Production and health data from the projects optiKuh and eMissionCow were used. Data consisted of weekly observations recorded over a 325-wk period in 2,387 GH and over a 300-wk period in 632 Fleckvieh (FV) primiparous and multiparous dairy cows from 13 dairy research farms in Germany. Variance and covariance components were estimated univariately or bivariately with linear random regression models for production data and threshold random regression models for health data. Heritabilities for DMI, EB, and REI were on average 0.17 and 0.15, 0.14 and 0.15, as well as 0.11 and 0.14 in GH and FV, respectively. Heritabilities on the underlying scale for mastitis, claw and leg diseases, and all diseases were on average 0.17 and 0.16, 0.18 and 0.12, as well as 0.15 and 0.11 in GH and FV, respectively. In GH, almost all genetic correlations were negative, especially in early lactation. Within the first 50 d in milk, genetic correlations between DMI and REI on the one hand and disease categories on the other ranged from -0.25 to -0.14 for mastitis, from -0.31 to -0.13 for claw and leg diseases, and from -0.58 to -0.30 for all diseases. Consequently, selection for lower DMI or REI could lead to a higher liability to diseases, especially in early lactation. A possibility to mitigate these undesirable side effects could be lactation stage-specific selection for FE. For FV, further studies with more data are needed to assess genetic relationships.

The molecular features of chronic lung allograft dysfunction in lung transplant airway mucosa.

Many lung transplants fail due to chronic lung allograft dysfunction (CLAD). We recently showed that transbronchial biopsies (TBBs) from CLAD patients manifest severe parenchymal injury and dedifferentiation, distinct from time-dependent changes. The present study explored time-selective and CLAD-selective transcripts in mucosal biopsies from the third bronchial bifurcation (3BMBs), compared to those in TBBs. We used genome-wide microarray measurements in 324 3BMBs to identify CLAD-selective changes as well as time-dependent changes and develop a CLAD classifier. CLAD-selective transcripts were identified with linear models for microarray data (limma) and were used to build an ensemble of 12 classifiers to predict CLAD. Hazard models and random forests were then used to predict the risk of graft loss using the CLAD classifier, transcript sets associated with rejection, injury, and time. T cell-mediated rejection and donor-specific antibody were increased in CLAD 3BMBs but most had no rejection. Like TBBs, 3BMBs showed a time-dependent increase in transcripts expressed in inflammatory cells that was not associated with CLAD or survival. Also like TBBs, the CLAD-selective transcripts in 3BMBs reflected severe parenchymal injury and dedifferentiation, not inflammation or rejection. While 3BMBs and TBBs did not overlap in their top 20 CLAD-selective transcripts, many CLAD-selective transcripts were significantly increased in both for example LOXL1, an enzyme controlling matrix remodeling. In Cox models for one-year survival, the 3BMB CLAD-selective transcripts and CLAD classifier predicted graft loss and correlated with CLAD stage. Many 3BMB CLAD-selective transcripts were also increased by injury in kidney transplants and correlated with decreased kidney survival, including LOXL1. Mucosal and transbronchial biopsies from CLAD patients reveal a diffuse molecular injury and dedifferentiation state that impacts prognosis and correlates with the physiologic disturbances. CLAD state in lung transplants shares features with failing kidney transplants, indicating elements shared by the injury responses of distressed organs.

Compensation-Based Cooperative MFAILC for Multiple Subway Trains Under Asynchronous Data Dropouts

This paper researches the cooperative control problem for multiple subway trains under the asynchronous data dropouts in both measurement channel and downlink channel. The compensation-based cooperative model free adaptive iterative learning control (cCMFAILC) for the multiple city subway trains (MCSTs) is proposed to avoid deterioration of the control performance due to data dropouts. First, the nonlinear subway train system is transformed into an equivalent dynamic linearization data model to describe the input-output dynamics of the subway train system. Next, the lost data is replaced by the corresponding data of the same time instant in the latest available iteration. And the cCMFAILC is designed to guarantee that the speed tracking errors of MCSTs are bounded along the iteration axis and the headway of neighboring subway trains is stabilized in a safe range. Finally, theoretical analysis and MCSTs simulations verify the validity of the proposed cCMFAILC scheme.

Data-Driven Formation Control for Unknown MIMO Nonlinear Discrete-Time Multi-Agent Systems With Sensor Fault.

A data-driven distributed formation control algorithm is proposed for an unknown heterogeneous non-affine nonlinear discrete-time MIMO multi-agent system (MAS) with sensor fault. For the considered unknown MAS, the dynamic linearization technique in model-free adaptive control (MFAC) theory is used to transform the unknown MAS into an equivalent virtual dynamic linearization data model. Then using the virtual data model, the structure of the distributed model-free adaptive controller is constructed. For the incorrect signal measurements due to the sensor fault, the radial basis function neural network (RBFNN) is first trained for the MAS under the fault-free case, then using the outputs of the well-trained RBFNN and the actual outputs of MAS under sensor fault case, the estimation laws of the unknown fault and system parameters in the virtual data model are designed with only the measured input-output (I/O) data information. Finally, the boundedness of the formation error is analyzed by the contraction mapping method and mathematical induction method. The effectiveness of the proposed algorithm is illustrated by simulation examples.

Time-series comparison of COVID-19 case fatality rates across 21 countries with adjustment for multiple covariates.

Although it is widely used as a measure for mortality, the case fatality rate (CFR) ofcoronavirus disease 2019 (COVID-19) can vary over time and fluctuate for many reasons otherthan viral characteristics. To compare the CFRs of different countries in equal measure, weestimated comparable CFRs after adjusting for multiple covariates and examined the mainfactors that contributed to variability in the CFRs among 21 countries. For statistical analysis, time-series cross-sectional data were collected from OurWorld in Data, CoVariants.org, and GISAID. Biweekly CFRs of COVID-19 were estimated bypooled generalized linear squares regression models for the panel data. Covariates includedthe predominant virus variant, reproduction rate, vaccination, national economic status,hospital beds, diabetes prevalence, and population share of individuals older than age 65. Intotal, 21 countries were eligible for analysis. Adjustment for covariates reduced variation in the CFRs of COVID-19 across countriesand over time. Regression results showed that the dominant spread of the Omicron variant,reproduction rate, and vaccination were associated with lower country-level CFRs, whereasage, the extreme poverty rate, and diabetes prevalence were associated with higher countrylevel CFRs. A direct comparison of crude CFRs among countries may be fallacious, especiallyin a cross-sectional analysis. Our study presents an adjusted comparison of CFRs over timefor a more proper comparison. In addition, our findings suggest that comparing CFRs amongdifferent countries without considering their context, such as the epidemic phase, medicalcapacity, surveillance strategy, and socio-demographic traits, should be avoided.

Dynamic event‐triggered model‐free adaptive security tracking control for constrained subway train system

AbstractIn this article, the dynamic event‐triggered model‐free adaptive security tracking control problem of the subway system with speed and traction/braking force constraints under aperiodic denial‐of‐service (DoS) attacks is studied. The complex subway train system is transformed into a linearization data model by using the dynamic linearization technique. For the subway system, both speed and traction/braking forces need to be restrained to ensure the safety of the subway equipment. In addition, the dynamic event‐triggered mechanism is designed to reduce communication frequency and prolong the hardware life cycle. At the same time, considering aperiodic DoS attacks in the network channel, the security tracking control problem is guaranteed. In the simulation, the algorithm is applied to the subway train system, and the effectiveness of this algorithm is verified with a comparison.

A split-and-conquer variable selection approach for high-dimensional general semiparametric models with massive data

Estimation and variable selection in partially linear models for massive data has been discussed by several authors. However, there does not seem to exist an established procedure for other semiparametric models, such as the semiparametric varying-coefficient linear model, the single index regression model, the partially linear errors-in-variables model, etc. In this paper, we propose a general procedure for variable selection in high-dimensional general semiparametric models by penalized semiparametric estimating equations. Under some regularity conditions, the oracle property is established, which the number of parameters is allowed to diverge. Furthermore, we also propose a split-and-conquer variable selection procedure for high-dimensional general semiparametric models with massive data. Under some weak regularity conditions, we establish the oracle property of the proposed procedure when the number of subsets does not grow too fast. What is more, the split-and-conquer procedure enjoys the oracle property as the penalized estimator by using all the dataset, and can substantially reduce computing time and computer memory requirements. The performance of the proposed method is illustrated via a real data application and numerical simulations.

Neural Network-Augmented Locally Adaptive Linear Regression Model for Tabular Data

Creating an interpretable model with high predictive performance is crucial in eXplainable AI (XAI) field. We introduce an interpretable neural network-based regression model for tabular data in this study. Our proposed model uses ordinary least squares (OLS) regression as a base-learner, and we re-update the parameters of our base-learner by using neural networks, which is a meta-learner in our proposed model. The meta-learner updates the regression coefficients using the confidence interval formula. We extensively compared our proposed model to other benchmark approaches on public datasets for regression task. The results showed that our proposed neural network-based interpretable model showed outperformed results compared to the benchmark models. We also applied our proposed model to the synthetic data to measure model interpretability, and we showed that our proposed model can explain the correlation between input and output variables by approximating the local linear function for each point. In addition, we trained our model on the economic data to discover the correlation between the central bank policy rate and inflation over time. As a result, it is drawn that the effect of central bank policy rates on inflation tends to strengthen during a recession and weaken during an expansion. We also performed the analysis on CO2 emission data, and our model discovered some interesting explanations between input and target variables, such as a parabolic relationship between CO2 emissions and gross national product (GNP). Finally, these experiments showed that our proposed neural network-based interpretable model could be applicable for many real-world applications where data type is tabular and explainable models are required.

Data-driven predictive point-to-point iterative learning control

Integrating the idea of predictive control and point-to-point iterative learning control, this paper presents a data-driven predictive point-to-point iterative learning control scheme for a class of unknown repetitive non-affine nonlinear SISO systems. The tracking task is driven by the optimal control input sequence generated by the proposed algorithm, and the tracking errors at the specified sampling time instants are minimized. The advantages of this scheme are that the structure of the controller and its stability analysis both are based on an equivalent dynamic linearization data model of the nonlinear system, and the proposed scheme does not involve the operation of matrix inversion. Numerical simulations verify the effectiveness of this method.

Revealing immune infiltrate characteristics and potential diagnostic value of immune-related genes in ulcerative colitis: An integrative genomic analysis.

Ulcerative colitis (UC) is an autoimmune disease of the colon. The aim of this study was to explore the characteristics of immune infiltrates in UC patients and identify immune-related diagnostic biomarkers for UC. Three gene expression profiles were acquired from the GEO database, followed by identification of differentially expressed genes (DEGs) by Linear Modeling of Microarray Data. Enrichment analysis of Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and Disease Ontology (DO) were performed to analyze the biological functions of DEGs. Subsequently, the single sample gene set enrichment analysis (ssGSEA) was performed to identify immune infiltration characteristics of UC. Correlations between diagnostic genes and immune infiltration were explored to identify markers with the greatest diagnostic potential, and a UC diagnostic model was subsequently constructed. Finally, the prediction performance of the model was quantified by nomogram, non-correlated nomogram, and ROC curve. A total of 3111 DEGs (1,608 up-regulated and 1,503 down-regulated genes) were identified. DEGs were significantly involved in the immune system and UC-related pathways. Immune infiltration profiles of colonic tissue were significantly different between healthy individuals and UC patients. High proportions of resting of aDCs, B cells, CD8+ T cells, DCs, iDCs, Macrophages, Neutrophils, pDCs, T helper cells, Tfh, Th1 cells, Th2 cells, TIL and Treg were found in UC samples. A 5-gene based diagnostic prediction model was constructed and the results of nomogram, non-correlated nomogram and ROC curve suggested the powerful diagnostic value of the model. This study identified the immune infiltrate characteristics and 5 immune-related genes for UC. The model based on the immune-related genes facilitates the early diagnosis of UC and provides a basis for the evaluation of the prognosis of UC.

Data-Driven Model-Free Adaptive Sliding Mode Control Based on FFDL for Electric Multiple Units

The electric multiple units (EMUs) have become a very convenient and powerful means of transportation in our daily life. Safe and punctual trajectory tracking control is the key to improve the performance of the EMUs system, but it is difficult to realize due to the influence of environmental uncertainty, coupling and nonlinearity. In this paper, a model-free adaptive sliding mode control (MFASMC) method is proposed for the EMUs. This method can solve the dependence of the model-based control method on the train model and eliminate the influence of external disturbances on the robust performance of the system. In this method, the running process of the EMUs is equivalent to a full format dynamic linearization (FFDL) data model, and a model-free adaptive controller (MFAC) is designed based on the data model. Then, to reduce the influence of measurement disturbance and improve the robustness of the system, a discrete sliding mode control (SMC) algorithm is introduced. Furthermore, to prevent the control input from being too large, the parameter estimation error is introduced as an additional correction term of the algorithm. In the end, the simulation experiment is carried out with CRH380A EMUs as the object. Compared with the traditional MFAC and the traditional SMC, the speed tracking effect of each power unit of the MFASMC algorithm is more effective, the change of control force is stable, the acceleration meets the requirements of driving, and has a strong inhibitory effect on external disturbances.

Model‐free adaptive and iterative learning composite control for subway train under actuator faults

AbstractIn this article, the model‐free adaptive and iterative learning composite control (CMFAC‐ILC) is proposed to ensure the speed and position tracking for the subway train system subjected to the iteration‐time‐varying actuator faults. First, a nonlinear subway train system is transformed into a faults‐related full‐form dynamic linearization data model (FFDLDM), which relies on the input, output, and faults data of the subway train system. The actuator faults and unknown nonlinear terms of the subway train system are estimated by the projection algorithm. Then, in the time domain, the model‐free adaptive control (MFAC) algorithm is utilized and unknown controller parameters are estimated by the RBFNN algorithm. In the iteration domain, a feedforward D‐type iterative learning control (ILC) algorithm is added to the outer loop of the MFAC algorithm. Finally, the theoretical analysis proves that the speed and position tracking errors of the subway train are bounded, the simulations demonstrate the effectiveness of the proposed composite control scheme of the subway train.

Predictive compensation based quantization iterative learning control for nonlinear nonaffine discrete‐time systems

AbstractIn this work, the problems of predictive compensation, unknown nonlinearity, and nonaffine structure are considered simultaneously for a quantized iterative learning control (QILC) design and analysis under a data‐driven framework. The compensation strategy can avoid deteriorated data transmission quality owing to limited channel capacities. First, a dynamic linearization methodology is employed to transform the nonlinear plant into a virtual iterative linear data model (iLDM) which includes all the input signals over a time‐window from the initial time instant to the current one. The iLDM is also used as a predictive model to estimate the unavailable information caused by the encoding–decoding mechanism. Then, a predictive compensation‐based QILC is proposed by optimizing quadratic functions, which includes an output prediction mechanism, a quantized iterative learning updating law, a quantized iterative parameter estimation law, and a resetting algorithm. The result is also extended to a class of MIMO nonlinear nonaffine discrete‐time systems. The developed control laws are data‐driven and independent of any system information. The theoretical results are proved by the use of contraction mapping principle and induction method. Examples are provided to verify the effectiveness of the proposed methods.