Search

Abstract Albatross Analytics is a statistical and data science data processing platform that researchers can use in disciplines of various fields. Albatross Analytics makes it easy to implement fundamental analysis for various regressions with random model effects, including Hierarchical Generalized Linear Models (HGLMs), Double Hierarchical Generalized Linear Models (DHGLMs), Multivariate Double Hierarchical Generalized Linear Models (MDHGLMs), Survival Analysis, Frailty Models, Support Vector Machines (SVMs), and Hierarchical Likelihood Structural Equation Models (HSEMs). We provide 94 types of dataset examples.

The diagnosis of a hazard can be classified into three key domains, particularly regarding the natural hazards, non-natural hazards and social hazards. The disasters which have actually happened in West Papua require considerable attention and consideration of the Indonesian Government, despite since they have handled as much as they can to provide solutions and make people feel secure and pleasant. The purpose of this study is to calculate the location-based social vulnerability in West Papua involves the components of Information, Technology, and Communication, Food Access, Natural Disaster, Social Protection Statement, Access to Financial Services, Description of the source of household income, Number of event floods, number of earthquake disasters, COVID-19 death cases, and Number of incidents of protest which are obtained from the National Socio-Economic Survey (SUSENAS) official statistics with the main focus of research on the millennial generation. After employ clustering of variables around latent variables with connectivity value of 3.9400794, Dunn 0.9373, and Silhouette 0.6333. Each factor provide a sign indicating a positive or negative effect on social vulnerability and finally a location cluster will be formed based on the index obtained.

As the sharing economy has emerged, the way customer perceives the service is shifting toward a combination of offline and online. The need for the service provider to understand its nature as well as the pertinent aspects regarding its characteristics is crucial. Previous research validated the influence of perceived online and offline service quality toward customer satisfaction and loyalty. However, with the distinctive dimensions of OFA service quality, its effects on customer satisfaction and the role of social innovativeness in satisfaction and loyalty linkage remain unexplored. Hence, this study attempts to investigate these relationships using the data obtained from customers of any OFA in Malaysia. Purposive sampling was employed and 227 collected responses were analyzed using variance-based partial least square path modeling. The results confirm the direct effect of online and offline service quality on customer loyalty and full mediation role of customer satisfaction. Besides, social innovativeness is found negatively moderates customer satisfaction and loyalty relationship. Implications and contributions of the study are also discussed.

The H-likelihood method proposed by Lee and Nelder (1996) is extensively used in a wide range of data. In terms of direction, repetitive measured data within classification can be examined employing hierarchical generalized linear models (HGLMs). Whether we are concerned in multiple endpoints which are correlated, instead Multivariate Double Hierarchical Generalized Linear Models (DHGLM) can be taken into consideration. This article addresses the implementation of this principle to vector selection and support machines. Based on the analysis with the fish morphology class Sardinella lemuru (Bali sardinella) and setting the best epsilon 0.7 cost 4 parameter reaching best performance: 0.2327401. Predictive value of fish sex was calculated 0.997319 and Region under the curve: 0.8967. At the same time, we extend the large-scale case studies for stress testing of the SVM method by using three datasets from UCI machine learning repository including the bank marketing dataset, the car evaluation database and human activity recognition using smartphones dataset. In a nutshell by employing SVM-DHGLM increased the accuracy, precision, recall, for feature selection and classification. Long story short, the $H$ -likelihood provides an excellent and usable structure for statistical inference of the unobservable general deterministic model, while preserving the advantages of the original probability structure for fixed parameters. We presume that more new groups of models will be created and that the $H$ -likelihood will be commonly used for their inferences and the application in big data and machine learning.

Construction signs alert drivers to the dangers of abnormally blocked roads. In the case of autonomous vehicles, construction signs should be detected automatically to prevent accidents. One might think that we can accomplish the goal easily using the popular deep-learning-based detectors, but it is not the case. To train the deep learning detectors to detect construction signs, we need a large amount of training images which contain construction signs. However, collecting training images including construction signs is very difficult in the real world because construction events do not occur frequently. To make matters worse, the construction signs might have dozens of different construction signs (i.e., contents). To address this problem, we propose a new method named content swapping. Our content swapping divides a construction sign into two parts: the board and the frame. Content swapping generates numerous synthetic construction signs by combining the board images (i.e., contents) taken from the in-domain images and the frames (i.e., geometric shapes) taken from the out-domain images. The generated synthetic construction signs are then added to the background road images via the cut-and-paste mechanism, increasing the number of training images. Furthermore, three fine-tuning methods regarding the region, size, and color of the construction signs are developed to make the generated training images look more realistic. To validate our approach, we applied our method to real-world images captured in South Korea. Finally, we achieve an average precision (AP50) score of 84.98%, which surpasses that of the off-the-shelf method by 9.15%. Full experimental results are available online as a supplemental video. The images used in the experiments are also released as a new dataset CSS138 for the benefit of the autonomous driving community.

The concentration of nitrogen dioxide (NO2) is worsening across the globe alongside growth in industrial and general anthropogenic activities. Due to its serious health implications with long-term exposure, studies on NO2 concentration have gained space in the academic literature. In this study, awareness is created on the levels of NO2 across four (4) locations within the Tema Metropolitan area, with specific interest in selecting locations and periods significantly saturated with NO2 within the study area. NO2 was measured using RKI Eagle, an instrument with a built-in sensor for a specific gas measurement. Measurements were taken day and night at sampling points around 100 meters apart in each location. Data collection was performed over a nine (9)-month period. The Generalized Linear model is explored for selecting locations and periods significantly affected by NO2. From the results, the fourth week (26th–31st) of July 2020, the fourth week (27th–31st) of December 2020, the first week (1st–7th) of January 2021, and the fourth week (24th–31st) of January 2021 recorded severe concentrations of NO2. Additionally, the lives of residents in the Oil Jetty and the VALVO hospital areas were found to be the most endangered, as they recorded significantly high concentrations of NO2. In a developing country such as Ghana, this study is useful for monitoring NO2 concentrations in similar areas to inform decision making and environmental policy formulation.

Dust levels around the Tema industrial area of the Greater Accra Region have seen no reduction in recent years. Even though at some periods in time a natural drop in dust pollution levels is assured, the overall variation characteristics of the concentration of PM2.5, PM10, and Total Suspended Particles (TSP) have not been studied in recent years. This paper examines the levels of dust pollution across four (4) locations within the Tema metropolitan area with a specific interest in selecting locations and periods (weeks) significantly affected by dust pollution within the study area. Data collection was done over a nine-month period using the Casella 712 Microdust Pro Kit equipment. Measurements were done day and night at sampling points about 100 m apart in a given location. Monitoring was conducted once a week during the day and at night with a sampling period of 24 h per location, for thirty-six weeks. The generalized linear models were explored in selecting locations and weeks significantly affected by dust pollution. The study results showed no significant difference between pollution levels across the four selected locations. Eight, eleven, and five weeks out of the 36 weeks recorded significantly high concentrations of PM2.5, PM10, and TSP respectively. In addition, two out of the selected four areas (the oil jetty area and the VALCO hospital area) were found to have significantly high concentrations of dust pollution. The study recommends that an urgent air quality control policy intervention be put in place to control the highly alarming levels of dust pollution concentrations to guarantee and protect human health within the study area and beyond.

Abstract We analyze a real data set pertaining to reindeer fecal pellet‐group counts obtained from a survey conducted in a forest area in northern Sweden. In the data set, over 70% of counts are zeros, and there is high spatial correlation. We use conditionally autoregressive random effects for modeling of spatial correlation in a Poisson generalized linear mixed model (GLMM), quasi‐Poisson hierarchical generalized linear model (HGLM), zero‐inflated Poisson (ZIP), and hurdle models. The quasi‐Poisson HGLM allows for both under‐ and overdispersion with excessive zeros, while the ZIP and hurdle models allow only for overdispersion. In analyzing the real data set, we see that the quasi‐Poisson HGLMs can perform better than the other commonly used models, for example, ordinary Poisson HGLMs, spatial ZIP, and spatial hurdle models, and that the underdispersed Poisson HGLMs with spatial correlation fit the reindeer data best. We develop R codes for fitting these models using a unified algorithm for the HGLMs. Spatial count response with an extremely high proportion of zeros, and underdispersion can be successfully modeled using the quasi‐Poisson HGLM with spatial random effects.

Covariate selection is a fundamental step when building sparse prediction models in order to avoid overfitting and to gain a better interpretation of the classifier without losing its predictive accuracy. In practice the LASSO regression of Tibshirani, which penalizes the likelihood of the model by the L1 norm of the regression coefficients, has become the gold-standard to reach these objectives. Recently Lee and Oh developed a novel random-effect covariate selection method called the modified unbounded penalty (MUB) regression, whose penalization function can equal minus infinity at 0 in order to produce very sparse models. We sought to compare the predictive accuracy and the number of covariates selected by these two methods in several high-dimensional datasets, consisting in genes expressions measured to predict response to chemotherapy in breast cancer patients. These comparisons were performed by building the Receiver Operating Characteristics (ROC) curves of the classifiers obtained with the selected genes and by comparing their area under the ROC curve (AUC) corrected for optimism using several variants of bootstrap internal validation and cross-validation. We found consistently in all datasets that the MUB penalization selected a remarkably smaller number of covariates than the LASSO while offering a similar-and encouraging-predictive accuracy. The models selected by the MUB were actually nested in the ones obtained with the LASSO. Similar findings were observed when comparing these results to those obtained in their first publication by other authors or when using the area under the Precision-Recall curve (AUCPR) as another measure of predictive performance. In conclusion, the MUB penalization seems therefore to be one of the best options when sparsity is required in high-dimension. Further investigation in other datasets is however required to validate these findings.

Motion analysis of the hyoid bone via videofluoroscopic study has been used in clinical research, but the classical manual tracking method is generally labor intensive and time consuming. Although some automatic tracking methods have been developed, masked points could not be tracked and smoothing and segmentation, which are necessary for functional motion analysis prior to registration, were not provided by the previous software. We developed software to track the hyoid bone motion semi-automatically. It works even in the situation where the hyoid bone is masked by the mandible and has been validated in dysphagia patients with stroke. In addition, we added the function of semi-automatic smoothing and segmentation. A total of 30 patients' data were used to develop the software, and data collected from 17 patients were used for validation, of which the trajectories of 8 patients were partly masked. Pearson correlation coefficients between the manual and automatic tracking are high and statistically significant (0.942 to 0.991, P-value

Recently a few articles have derived robust first-order rotatable and D-optimal designs for the lifetime response having distributions gamma, lognormal, Weibull, exponential assuming errors that are correlated with different correlation structures such as autocorrelated, intra-class, inter-class, tri-diagonal, compound symmetry. Practically, a first-order model is an adequate approximation to the true surface in a small region of the explanatory variables. A second-order model is always appropriate for an unknown region, or if there is any curvature in the system. The current article aims to extend the ideas of these articles for second-order models. Invariant (free of the above four distributions) robust (free of correlation parameter values) second-order rotatable designs have been derived for the intra-class and inter-class correlated error structures. Second-order rotatability conditions have been derived herein assuming the response follows non-normal distribution (any one of the above four distributions) and errors have a general correlated error structure. These conditions are further simplified under intra-class and inter-class correlated error structures, and second-order rotatable designs are developed under these two structures for the response having anyone of the above four distributions. It is derived herein that robust second-order rotatable designs depend on the respective error variance covariance structure but they are independent of the correlation parameter values, as well as the considered four response lifetime distributions. [ABSTRACT FROM AUTHOR]

Summary: We study properties of the maximum h‐likelihood estimators for random effects in clustered data. To define optimality in random effects predictions, several foundational concepts of statistics such as likelihood, unbiasedness, consistency, confidence distribution and the Cramer–Rao lower bound are extended. Exact probability statements about interval estimators for random effects can be made asymptotically without a prior assumption. Using the binary‐matched pair example, we illustrated that the use of random effects recover information, leading to the boon on estimating treatment effects. [ABSTRACT FROM AUTHOR]

The determinants of FEV1 and FVC of 470 primary lung cancer patients with 17 characters who underwent major lung resections have been derived in the report. The mean FEV1 is positively associated with FVC (p<0.0001), and it is negatively associated with performance status at level 2 (p=0.0092), level 3 (p=0.0196), while its variance is positively associated with haemoptysis (p<0.0001), dyspnoea (p<0.0001), and it is negatively associated with FVC (p<0.0001), age (p<0.0001), performance status at level 2 (p<0.0001), level 3 (p<0.0001). On the other hand, mean FVC is negatively associated with age (P<0.0001), performance status at level 2 (p=0.0431), haemoptysis (p=0.0351), diabetes mellitus (p=0.0276), asthma (p<0.0001), while it is positively associated with diagnosis (p=0.0032) and cough (p=0.0676). Younger lung cancer patients with no advanced diseases such as asthma, diabetes mellitus, haemoptysis have higher FVC or FEV1. Care should be taken for advanced disease lung cancer patients during lung resection. [ABSTRACT FROM AUTHOR]

The glucose metabolisms and serum lipids are assumed as possible intermediary mechanisms in linking breast cancer (BC) and obesity. The current report examines the associations between diabetes mellitus (DM) markers (glucose and insulin) and BC markers (monocyte chemoattractant protein-1 (MCP-1), resistin, adiponectin, leptin). The glucose model shows that mean glucose levels are higher for breast cancer women (p=0.0222) then normal. Mean glucose levels are positively associated with leptin (p<0.0001) and homeostasis model assessment score insulin resistance (HOMA-IR) (p<0.0001), while they are negatively associated with interaction effects HOMA-IR*leptin (p<0.0001) and leptin*adiponectin (p=0.0883). On the other hand, variance of glucose levels is positively associated with HOMA-IR (p<0.0001) and resistin (p=0.0218), while it is negatively associated with leptin (p<0.0001), MCP-1 (p=0.0115). Insulin model shows that mean insulin levels are positively associated with HOM-IR (p<0.0001), leptin (p=0.0009), age*MCP-1 (p=0.0909), glucose*adiponectin (p=0.0424), glucose*resistin (p<0.0001), HOMAIR* MCP-1 (p<0.0001), while they are negatively associated with MCP-1 (p=0.0264), resistin (p<0.0001), adiponectin(p=0.0783), glucose*HOMAIR (p<0.0001), leptin*adiponectin (p=0.0713). The variance of insulin levels is higher for breast cancer women (p=0.0003) than normal. Again, it is positively associated with MCP-1 (p=0.0014), HOMA-IR (p<0.0001), while it is negatively associated with leptin (p=0.0828) and glucose*MCP-1 (p=0.0003). Many more relationships between BC and DM markers are also reported in the current article. It is concluded that both DM and BC markers have very complex closely interlinked relationships. [ABSTRACT FROM AUTHOR]

Breast cancer is in most cases noticed over the world as the second cause of cancer-allied death in women. Monocyte chemoattractant protein-1 (MCP-1) plays an essential role in the opening and advancement of cancer. The report derives the association of MCP-1 with age, BMI, insulin, glucose and other breast cancer biomarkers such as leptin, resistin, adiponectin, HOMA based on a real data set. It is obtained herein that mean MCP-1 is positively associated with BMI (p<0.0001), resistin (p<0.0001) and the interaction effect of insulin and leptin (i.e., insulin*leptin) (p<0.0001), while it is negatively associated with insulin (p<0.0001) and leptin (p<0.0001). The variance of MCP-1 is positively associated with age*insulin (p=0.0025), leptin*resistin (p=0.0176) and glucose*leptin (p=0.0819), while it is negatively associated with age (p=0.0706), homeostasis model assessment score (HOMA) (p=0.0055), leptin (p=0.0198) and resistin (p=0.0777). Only three partially significant effects (approximately 8% level of significance) are included in the Gamma variance model, while they are significant in the Log-normal variance model. These are accounted as a confounder in the model in view of Epidemiology. It is concluded herein that MCP-1 is higher for women with high BMI, higher levels of resistin, low levels of both insulin and leptin, along with their high interaction effect. [ABSTRACT FROM AUTHOR]

Background and objectives: The ubiquitous Internet connections by smartphones weakened the traditional boundaries between computers and mobile phones. We sought to explore whether smartphone-related problems differ from those of computer use according to gender using latent class analysis (LCA). Methods: After informed consents, 555 Korean middle-school students completed surveys on gaming, Internet use, and smartphone usage patterns. They also completed various psychosocial instruments. LCA was performed for the whole group and by gender. In addition to ANOVA and X² tests, post-hoc tests were conducted to examine differences among the LCA subgroups. Results: In the whole group (n = 555), four subtypes were identified: dual-problem users (49.5%), problematic Internet users (7.7%), problematic smartphone users (32.1%), and "healthy" users (10.6%). Dual-problem users scored highest for addictive behaviors and other psychopathologies. The gender-stratified LCA revealed three subtypes for each gender. With dual-problem and healthy subgroup as common, problematic Internet subgroup was classified in the males, whereas problematic smartphone subgroup was classified in the females in the gender-stratified LCA. Thus, distinct patterns were observed according to gender with higher proportion of dual-problem present in males. While gaming was associated with problematic Internet use in males, aggression and impulsivity demonstrated associations with problematic smartphone use in females. Conclusions: An increase in the number of digital media-related problems was associated with worse outcomes in various psychosocial scales. Gaming may play a crucial role in males solely displaying Internet-related problems. The heightened impulsivity and aggression seen in our female problematic smartphone users requires further research. [ABSTRACT FROM AUTHOR]

Because bioequivalence studies are performed using a crossover design, information on the intrasubject coefficient of variation (intra-CV) for pharmacokinetic measures is needed when determining the sample size. However, calculated intra-CVs based on bioequivalence results of identical generic drugs produce different estimates. In this study, we collected bioequivalence results using public resources from the Ministry of Food and Drug Safety (MFDS) and calculated the intra-CVs of various generics. For the generics with multiple bioequivalence results, pooled intra-CVs were calculated. The estimated intra-CVs of 142 bioequivalence studies were 14.7±8.2% for AUC and 21.7±8.8% for Cmax. Intra-CVs of Cmax were larger than those of area under the concentrationtime curve (AUC) in 129 studies (90.8%). For the 26 generics with multiple bioequivalence results, the coefficients of variation of intra-CVs between identical generics (mean±sd (min ~ max)) were 38.0±24.4% (1.9 ~ 105.3%) for AUC and 27.9±18.2% (4.0 ~ 70.1%) for Cmax. These results suggest that substantial variation exists among the bioequivalence results of identical generics. In this study, we presented the intra-CVs of various generics with their pooled intra-CVs. The estimated intra-CVs calculated in this study will provide useful information for planning future bioequivalence studies. (This is republication of the article 'Transl Clin Pharmacol 2017;25:179-182' retracted from critical typographic errors. See the 'Retraction and Republication section of this issue for further information). [ABSTRACT FROM AUTHOR]

Current assessment of gene-gene interactions is typically based on separate parallel analysis, where each interaction term is tested separately, while less attention has been paid on simultaneous estimation of interaction terms in a prediction model. As the number of interaction terms grows fast, sparse estimation is desirable from statistical and interpretability reasons. There is a large literature on sparse estimation, but there is a natural hierarchy between the interaction and its corresponding main effects that requires special considerations. We describe random-effect models that impose sparse estimation of interactions under both strong and weak-hierarchy constraints. We develop an estimation procedure based on the hierarchical-likelihood argument and show that the modelling approach is equivalent to a penalty-based method, with the advantage of the models being more transparent and flexible. We compare the procedure with some standard methods in a simulation study and illustrate its application in an analysis of gene-gene interaction model to predict body-mass index. [ABSTRACT FROM AUTHOR]

Background: Controlling the false discovery rate is important when testing multiple hypotheses. To enhance the detection capability of a false discovery rate control test, we applied the likelihood ratio-based multiple testing method in neuroimage data and compared the performance with the existing methods. Methods: We analysed the performance of the likelihood ratio-based false discovery rate method using simulation data generated under independent assumption, and positron emission tomography data of Alzheimer's disease and questionable dementia. We investigated how well the method detects extensive hypometabolic regions and compared the results to those of the conventional Benjamini Hochberg-false discovery rate method. Results: Our findings show that the likelihood ratio-based false discovery rate method can control the false discovery rate, giving the smallest false non-discovery rate (for a one-sided test) or the smallest expected number of false assignments (for a two-sided test). Even though we assumed independence among voxels, the likelihood ratio-based false discovery rate method detected more extensive hypometabolic regions in 22 patients with Alzheimer's disease, as compared to the 44 normal controls, than did the Benjamini Hochberg-false discovery rate method. The contingency and distribution patterns were consistent with those of previous studies. In 24 questionable dementia patients, the proposed likelihood ratio-based false discovery rate method was able to detect hypometabolism in the medial temporal region. Conclusions: This study showed that the proposed likelihood ratio-based false discovery rate method efficiently identifies extensive hypometabolic regions owing to its increased detection capability and ability to control the false discovery rate. [ABSTRACT FROM AUTHOR]

We present the frailtyHL package for fitting semi-parametric frailty models using hlikelihood. This package allows lognormal or gamma frailties for random-effect distribution, and it fits shared or multilevel frailty models for correlated survival data. Functions are provided to format and summarize the frailtyHL results. The estimates of fixed effects and frailty parameters and their standard errors are calculated. We illustrate the use of our package with three wellknown data sets and compare our results with various alternative R-procedures. [ABSTRACT FROM AUTHOR]

Copyright of Applied Stochastic Models in Business & Industry is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Purpose - With a team interaction analysis model, the authors sought to identify a varying range of individual and collective intellectual behaviors in a series of communicative intents particularly expressed with multimodal interaction methods. In this paper, the authors aim to present a new construct (i.e. collective intelligence ratio (CIR)) which refers to a numeric indicator representing the degree of intelligence of a team in which each team member demonstrates an individual intelligence ratio (IR) specific to a team goal. Design/methodology/approach - The authors analyzed multimodal team interaction data linked to communicative intents with a Poisson-hierarchical generalized linear model (HGLM). Findings - The study found evidence of a distinctive IR for each team member in selecting a communicative method for a certain task, ultimately leading to varying degrees of team CIR. Research limitations/implications - The authors limited the type and nature of human intelligence observed with a very short list of categories. Also, the data were evaluated by only one subject matter expert, leading to reliability issues. Therefore, generalization should be limited to situations in which teams, with pre-specified team goals and tasks, are collaborating in multimodal interaction environments. Practical implications - This study presents potential ways to directly or indirectly optimize team performance by identifying and incorporating IRs and CIRs in team composition strategies. Originality/value - In the literature of team cognition and performance, the authors offer a new insight on team schema by suggesting a new task-expertise-person (TEP) unit integrating information on who uses what communicative methods to best tackle on what cognitive task (i.e. optimum cognition with least cognitive burden). Individual and collective intelligence ratios should be considered as new extensions to conventional transactive memory systems in multimodal team interaction scenarios. [ABSTRACT FROM AUTHOR]

Canonical covariance analysis (CCA) has gained popularity as a method for the analysis of two sets of high-dimensional genomic data. However, it is often difficult to interpret the results because canonical vectors are linear combinations of all variables, and the coefficients are typically nonzero. Several sparse CCA methods have recently been proposed for reducing the number of nonzero coefficients, but these existing methods are not satisfactory because they still give too many nonzero coefficients. In this paper, we propose a new random-effect model approach for sparse CCA; the proposed algorithm can adapt arbitrary penalty functions to CCA without much computational demands. Through simulation studies, we compare various penalty functions in terms of the performance of correct model identification. We also develop an extension of sparse CCA to address more than two sets of variables on the same set of observations. We illustrate the method with an analysis of the NCI cancer dataset. [ABSTRACT FROM AUTHOR]

Frailty models with a non-parametric baseline hazard are widely used for the analysis of survival data. However, their maximum likelihood estimators can be substantially biased in finite samples, because the number of nuisance parameters associated with the baseline hazard increases with the sample size. The penalized partial likelihood based on a first-order Laplace approximation still has non-negligible bias. However, the second-order Laplace approximation to a modified marginal likelihood for a bias reduction is infeasible because of the presence of too many complicated terms. In this article, we find adequate modifications of these likelihood-based methods by using the hierarchical likelihood. [ABSTRACT FROM AUTHOR]

We propose a multivariate volatility model for the behaviour of eight international equity indices. We show that many volatility models with heavy tails in financial work can be viewed as the GLM class of models with random effects in the dispersion. Hence, the h-likelihood approach, which provides efficient and simpler algorithms for GLM class, can be used as an estimation method for models used in finance. A comparison of the h-likelihood estimators with the ML estimators is made and its relative merits are discussed. [ABSTRACT FROM AUTHOR]

This article presents research on spatial and spatio-temporal modelling. In particular, the article examines a comparison of the hierarchical likelihood and Bayesian methods to spatial epidemiological modelling. Three examples are investigated in the research--low birth weight, and cancer mortality in the state of South Carolina and lip cancer statistics from Scotland--with results showing that mean estimates from the hierarchical and Bayesian methods are nearly identical. Further implications of the research are addressed.

Generalized linear models (GLMs) are widely used for data analysis; however, their maximum likelihood estimators can be sensitive to outliers. We propose new statistical models that allow robust inferences from the GLM class of models, including Poisson and binomial GLMs, and their extension to generalized linear mixed models. The likelihood score equations from the new models give estimators with bounded influence, so that the resulting estimators are robust against outliers while maintaining high efficiency in the absence of outliers. [ABSTRACT FROM AUTHOR]

Abstract  In recent issues of this journal it has been asserted in two papers that the use of h-likelihood is wrong, in the sense of giving unsatisfactory estimates of some parameters for binary data (Kuk and Cheng, 1999; Waddington and Thompson, 2004) or theoretically unsound (Kuk and Cheng, 1999). We wish to refute both these assertions. [ABSTRACT FROM AUTHOR]

This letter proposes a wavelet denoising method in the presence of missing data. This approach is based on a coupling of wavelet shrinkage and hierarchical (or h)-likelihood method. The h-likelihood provides an effective imputation methodology of missing data to give wavelet estimators for signals and motivates a fast and simple algorithm. The method can be easily extended to other settings, such as image denoising. Simulation studies demonstrate empirical properties of the proposed method. [ABSTRACT FROM AUTHOR]

We propose a class of double hierarchical generalized linear models in which random effects can be specified for both the mean and dispersion. Heteroscedasticity between clusters can be modelled by introducing random effects in the dispersion model, as is heterogeneity between clusters in the mean model. This class will, among other things, enable models with heavy-tailed distributions to be explored, providing robust estimation against outliers. The h-likelihood provides a unified framework for this new class of models and gives a single algorithm for fitting all members of the class. This algorithm does not require quadrature or prior probabilities. [ABSTRACT FROM AUTHOR]

When there are two alternative random-effect models leading to the same marginal model, inferences from one model can be used for the other model. We illustrate how a likelihood method for fitting models with independent random effects can be applied to seemingly very different models with correlated random effects. We also discuss some merits of using these alternative models. [ABSTRACT FROM AUTHOR]