Search

Chronic obstructive pulmonary disease (COPD) is a preventable, treatable, progressive chronic disease characterized by persistent airflow limitation. Patients with COPD deserve special consideration regarding treatment in this fragile population for preclinical health management. Therefore, this paper proposes a novel lung radiomics combination vector generated by a generalized linear model (GLM) and Lasso algorithm for COPD stage classification based on an auto-metric graph neural network (AMGNN) with a meta-learning strategy. Firstly, the parenchyma images were segmented from chest high-resolution computed tomography (HRCT) images by ResU-Net. Second, lung radiomics features are extracted from the parenchyma images by PyRadiomics. Third, a novel lung radiomics combination vector (3 + 106) is constructed by the GLM and Lasso algorithm for determining the radiomics risk factors (K = 3) and radiomics node features (d = 106). Last, the COPD stage is classified based on the AMGNN. The results show that compared with the convolutional neural networks and machine learning models, the AMGNN based on constructed novel lung radiomics combination vector performs best, achieving an accuracy of 0.943, precision of 0.946, recall of 0.943, F1-score of 0.943, and ACU of 0.984. Furthermore, it is found that our method is effective for COPD stage classification.

Categorical data derived from qualitative classifications or countable quantitative data are common in biological scientific work and crop breeding. Categorical data analyses are important for drawing correct inferences from experiments. However, categorical data can introduce unique issues in data analysis. This paper discusses common problems arising from categorical variable analysis and modeling, demonstrates the issues or risks of misapplying analysis, and suggests approaches to address data analysis challenges using two data sets from alfalfa breeding programs. For each data set, we present several analysis methods, e.g., simple t-test, analysis of variance (ANOVA), split plot analysis, generalized linear model (glm), generalized linear mixed model (glmm) using R with R markdown, and with the standard statistical analysis software SAS/JMP. The goal is to demonstrate good analysis practices for categorical data by comparing the potential ‘bad’ analyses with better ones, avoiding too much reliance on reaching a significant p-value of 0.05, and navigating the morass of ever-increasing numbers of potential R functions. The three main aspects of this research focus on choosing the right data distribution to use, using the correct error terms for hypothesis test p-values including the right type of sum of the squares (Type I, II, and III), and proper statistical models for categorical data analysis. Our results show the importance of good statistical analysis practice to help agronomists, breeders, and other researchers apply appropriate statistical approaches to draw more accurate conclusions from their data.

Nowadays, one of the biggest challenges in banking sector, certainly, is assessment of the client’s creditworthiness. In order to improve the decision-making process and risk management, banks resort to using data mining techniques for hidden patterns recognition within a wide data. The main objective of this study is to build a high-performance customized credit scoring model. The model named Reliable client is based on Bank’s real dataset and originally built by applying four different classification algorithms: decision tree (DT), naive Bayes (NB), generalized linear model (GLM) and support vector machine (SVM). Since it showed the greatest results, but also seemed as the most appropriate algorithm, the adopted model is based on GLM algorithm. The results of this model are presented based on many performance measures that showed great predictive confidence and accuracy, but we also demonstrated significant impact of data pre-processing on model performance. Statistical analysis of the model identified the most significant parameters on the model outcome. In the end, created credit scoring model was evaluated using another set of real data of the same Bank.

Purpose: This empirical research aims to predict the distinguishing variables of passengers who did or did not survive while traveling in the famous Titanic liner, which sunk in 1912. Design/Methodology/Approach: The binary logistic regression analysis empirically analyzes the secondary dataset available for 1046 passengers. Variables such as passenger’s gender, age, family composition, ticket class, number of parents with/without children, and number of siblings and/or spouses were opted to examine the differences between the binary dependent variable (Passenger Survived/ Not Survived). Findings: The study results indicate that all the variables are statistically significant in the model, with passenger's gender being the most significant predictor followed by passenger’s ticket class. The survival chances of passengers decreased for male passengers compared to their counterparts (female passengers) for the sample data [Exp(β)=0.080], for the passengers of age more than 21 years compared to passengers of age less than and equal to 21 years [Exp(β)=0.576], and for passengers with ticket class second and third compared to first-class ticket holders [Exp(β)=0.412]. In contrast, there was a greater chance of survival for families traveling together with parents, siblings, spouses compared to single travelers [Exp(β)=1.823]. Implications/Originality/Value: The study is a classic example of the application of binary logistic regression analysis using EVIEWS software.

The cross-border movement of capital has suffered due to the COVID-19 pandemic since December 2019. Nevertheless, it is unrealistic for multinational companies to withdraw giant global value chains (GVCs) overnight because of the pandemic. Instead, active discussions and achievements of deals in cross-border mergers and acquisitions (M&amp, As) are expected in the post-COVID-19 era among various other market entry modes, considering the growing demand in high technologies in societies. This paper analyzes particular determinants of cross-border mergers and acquisitions (M&amp, As) during the pandemic year (2020) based on cross-sectional datasets by employing quasi-Poisson and negative binomial regression models. According to the empirical evidence, COVID-19 indices do not hamper M&amp, A deals in general. This indicates that managerial capabilities of the coronavirus, not the outbreak itself, determined locational decisions of M&amp, A deals during the pandemic. In this vein, it is expected that the vaccination rate will become a key factor of locational decision for M&amp, A deals in the near future. Furthermore, countries that have been outstanding in coping with COVID-19 and thus serve as a good example for other nations may seize more opportunities to take a leap forward. In addition, as hypothesized, the results present positive and significant associations with M&amp, A deals and the SDG index, confirming the resource-based theory of internationalization. In particular, the achievement of SDGs seems to exercise much influence in developing countries for M&amp, A bidders during the pandemic year. This indicates that the pandemic demands a new zeitgeist that pursues growth while resolving existing but disregarded environmental issues and cherishes humanitarian values, for all countries, non-exceptionally, standing at the start line of the post-COVID-19 era.

Conventional EEG-fMRI methods have been proven to be of limited use in the sense that they cannot reveal the information existing in between the spikes. To resolve this issue, the current study obtains the time series of the epileptic components detected on EEG and uses them to fit the Generalized Linear Model (GLM), as a substitution for classical regressors. This approach allows for more precise localization, and equally importantly, the prediction of the future behavior of the epileptic generators. The proposed method approaches the localization process in the source domain (component), rather than the electrode domain (EEG), and localizes the generators through investigating the spatial correlation between the spike template and the candidate components, as well as the medical records of the patient. To evaluate the contribution of EEG-fMRI and concordance between EEG and fMRI, this method was performed on 30 patients with refractory epilepsy. The results demonstrated the significant numbers of 29 and 24 for concordance and contribution, respectively, which marks a rise as compared to the existing literature. This study also shows that while conventional methods often fail to properly localize the epileptogenic zones that are located in deep brain structures, the proposed method can be of particular use. For further evaluation, the concordance level between SOZ and IED-related BOLD clusters has been quantitatively investigated for different brain structures. Therefore, the distance between the BOLD clusters and IED/SOZ locations was measured in all patients. The results showed the superiority of the proposed method in delineating the spike-generating network compared to more conventional EEG-fMRI approaches. In all, the proposed method goes beyond the conventional methods by breaking the dependency on inside-the-scanner spikes and using the outside-the-scanner spike templates and the selected components, instead; achieving an accuracy of as high as 97%. Doing so, this method contributes to improving the yield of EEG-fMRI and creates a more realistic perception of the neural behavior of epileptic generators which is almost without precedent in the literature.

Perfluoroalkyl and polyfluoroalkyl substances (PFASs) are widely used in consumer products. However, the role of PFAS in infertility is still poorly understood. A total of 788 women from the 2013–2016 nationally representative NHANES were included to explore the association between PFAS exposure and self-reported infertility. Six PFAS, including PFDE, PFNA, PFHxS, n-PFOA, n-PFOS, and Sm-PFOS, were detected by online SPE-HPLC-TIS-MS/MS. We used the generalized linear regression model (GLM), generalized additive models (GAM), and Bayesian kernel machine regression (BKMR) to assess the single effects, non-linear relationships, and mixed effects on women’s infertility, respectively. The prevalence of self-reported infertility was 15.54% in this study. In GLM, n-PFOA showed a negative association with self-reported infertility in women for the Q3 (OR: 0.396, 95% CI: 0.119, 0.788) and Q4 (OR: 0.380, 95% CI: 0.172–0.842) compared with Q1 (p for trend = 0.013). A negative trend was also observed in n-PFOS and ∑PFOS (p for trend < 0.05). In GAM, a non-linear relationship was revealed in Sm-PFOS, which exhibits a U-shaped relationship. The BKMR model indicated that there might be a joint effect between PFAS and women’s infertility, to which PFNA contributed the highest effect (PIP = 0.435). Moreover, age stratification analysis showed a different dose–response curve in under and above 35 years old. Women under the age of 35 have a more noticeable U-shaped relationship with infertility. Therefore, the relatively low level of mixed PFAS exposure was negatively associated with self-reported infertility in women in general, and the impact of PFAS on infertility may vary among women of different age groups. Further studies are needed to determine the etiological relationship.

Although gambling is forbidden for minors, the prevalence of gambling among adolescents is increasing. In order to improve preventive interventions, more evidence on predictors of gambling onset is needed. A longitudinal study was proposed to (1) establish the prevalence of gambling, (2) identify factors associated with gambling behavior the following year, and (3) adjust a model to predict gambling behavior. A cohort of 1074 students (13&ndash, 18 years old) was followed for 12 months. The prevalence of gambling reached 42.0% in the second measure. Boys gambled 2.7 times more than girls, and the highest percentages of gambling onset showed up between 13 and 14 years old. Gambling onset and maintenance was associated with gender, age, sensation-seeking, risk perception, self-efficacy for not gambling, parents&rsquo, attitude towards gambling, group pressure (friends), subjective norm, exposure to advertising, accessibility, normative perception, gambling in T1 and parents gambling behavior. Gender, gambling in T1 and risk perception were significant in all three logistic adjusted regression models, with the fourth variable being sensation seeking, peer pressure (friends) and accessibility, respectively. It is suggested that universal prevention should be aimed preferably at children under 15 years old and to alert regulators and public administrations to the directly proportional relationship between accessibility and gambling onset.

The paper authenticated the need for separate positive integer time series model(s). This was done from the standpoint of a proposal for both mixtures of continuous and discrete time series models. Positive integer time series data are time series data subjected to a number of events per constant interval of time that relatedly fits into the analogy of conditional mean and variance which depends on immediate past observations. This includes dependency among observations that can be best described by Generalized Autoregressive Conditional Heteroscedasticity (GARCH) model with Poisson distributed error term due to its positive integer defined range of values. As a result, an integer GARCH model with Poisson distributed error term was formed in this paper and called Integer Generalized Autoregressive Conditional Heteroscedasticity (INGARCH). Iterative Reweighted Least Square (IRLS) parameter estimation technique type of the Generalized Linear Models (GLM) was adopted to estimate parameters of the two spilt models; Linear and Log-linear INGARCH models deduced from the identity link function and logarithmic link function, respectively. This resulted from the log-likelihood function generated from the GLM via the random component that follows a Poisson distribution. A study of monthly successful bids of auction from 2003 to 2015 was carried out. The Probabilistic Integral Transformation (PIT) and scoring rule pinpointed the uniformity of the linear INGARCH than that of the log-linear INGARCH in describing first order autocorrelation, serial dependence and positive conditional effects among covariates based on the immediate past. The linear INGARCH model outperformed the log-linear INGARCH model with (AIC = 10514.47, BIC = 10545.01, QIC = 34128.56) and (AIC = 37588.83, BIC = 37614.28, QIC = 37587.3), respectively.

Pricing of repair and maintenance (R&M) contracts is one among the most important processes carried out at Scania. Predictions of repair costs at Scania are carried out using experience-based prediction methods which do not involve statistical methods for the computation of average repair costs for contracts terminated in the recent past. This method is difficult to apply for a reference population of rigid Scania trucks. Hence, the purpose of this study is to perform suitable statistical modelling to predict repair costs of four variants of rigid Scania trucks. The study gathers repair data from multiple sources and performs feature selection using the Akaike Information Criterion (AIC) to extract the most significant features that influence repair costs corresponding to each truck variant. The study proved to show that the inclusion of operational features as a factor could further influence the pricing of contracts. The hurdle Gamma model, which is widely used to handle zero inflations in Generalized Linear Models (GLMs), is used to train the data which consists of numerous zero and non-zero values. Due to the inherent hierarchical structure within the data expressed by individual chassis, a hierarchical hurdle Gamma model is also implemented. These two statistical models are found to perform much better than the experience-based prediction method. This evaluation is done using the mean absolute error (MAE) and root mean square error (RMSE) statistics. A final model comparison is conducted using the AIC to draw conclusions based on the goodness of fit and predictive performance of the two statistical models. On assessing the models using these statistics, the hierarchical hurdle Gamma model was found to perform predictions the best

Soils are a vital part of the natural environment and one of the most important natural resources. The European Mediterranean Regions are particularly sensitive to soil degradation. Especially in Italy many landscapes are prone to water related soil erosion processes such as rill-interrill erosion (sheet erosion), gullies and complex processes forming “calanchi” corresponding to the English term “Badlands”. Calanchi develop preferentially on Plio-Pleistocene deposits, which are highly susceptible to aquatic erosion processes. This study focusses on the assessment of rill-interrill erosion as well as calanchi erosion forms and features in previously none or very little-studied areas of the Langhe-Roero and Monferrato, located in the Liguria and Piedmont Region, Northern Apennines, Italy. In this study we assess the driving factors and the spatial distribution of rill-interrill and calanchi erosion processes using a geostatistical/geostochastic modelling framework. Therefore, a Maximum Entropy Model, a Generalized Linear Model and a Boosted Regression Tree approach were applied. As independent environmental variables we selected DEM based morphometric information as well as pedologic, and geologic input data. Moreover, we explore remote sensing techniques to derive information on the vegetation density and vitality. The stochastic models applied show, that the two very different soil erosion phenomena can be differentiated well with the chosen stochastic techniques. The main drivers controlling the spatial distribution of aquatic soil erosion in the study area are soil type, slope, elevation and topographic wetness index. The results of the MaxEnt and BRT model were confirmed by the GLM application in terms of the significance of predictor variables. Generally, the models show an acceptable to excellent performance with BRT and GLM outperforming MaxEnt. The single independent variable response curves gave valuable insights into the process triggering the erosion forms. We were able to clearly differentiate between the processes looking at the response curves that show reasonable characteristics with all applied models.

Single neurons can dynamically change the gain of their spiking responses to account for shifts in stimulus variance. Moreover, gain adaptation can occur across multiple timescales. Here, we examine the ability of a simple statistical model of spike trains, the generalized linear model (GLM), to account for these adaptive effects. The GLM describes spiking as a Poisson process whose rate depends on a linear combination of the stimulus and recent spike history. The GLM successfully replicates gain scaling observed in Hodgkin-Huxley simulations of cortical neurons that occurs when the ratio of spike-generating potassium and sodium conductances approaches one. Gain scaling in the GLM depends on the length and shape of the spike history filter. Additionally, the GLM captures adaptation that occurs over multiple timescales as a fractional derivative of the stimulus variance, which has been observed in neurons that include long timescale after hyperpolarization conductances. Fractional differentiation in GLMs requires long spike history that span several seconds. Together, these results demonstrate that the GLM provides a tractable statistical approach for examining single-neuron adaptive computations in response to changes in stimulus variance.

Affecting daily lives of millions of people, Epilepsy is a common central nervous system (neurological) disorder where cell activity in brain is disturbed, causing recurrent seizures. Epilepsy can be treated commonly by medications. Be that as it may, medications do not always work as well as one may have hoped, and thus, some patients tend to resort to surgeries. The primary challenge in such surgeries, and by extension any other surgery where some part of brain may need to be disabled, disconnected or removed, is managing to pose no threat to the critical healthy textures adjacent or close to the part being operated on. Therefore, the precise localization of epileptic focus is a matter of vital importance in treating this condition. Various algorithms have been proposed to localize the brain sources and thus to determine the epileptic focus, however, none has yet been able to offer a solution to effectively address this issue. With EEG signal containing temporal information and fMRI carrying spatial information, it is hoped that the combination of the two can yield optimal results. In this case study, we first remove the artifacts caused by EEG gradients, and proceed to study the signal in and outside the scanner by localizing the brain sources. The simultaneous processing of EEG-fMRI enables us to make use of the temporal information in EEG to analyze fMRI. Epileptic foci are finally localized based on GLM method. This study has been conducted on 2 medication-resistant patients with epilepsy whose data was recorded in Iran National Brain Mapping Centre. The results suggest a significant improvement in localization accuracy compared to existing methods in the literature. Keywords:Simultaneous EEG-fMRI; Epilepsy; Independent Component Analysis (ICA); Blood-oxygen-level dependent imaging (BOLD); Generalized Linear Model (GLM)

Background: Hi-C and its variant techniques have been developed to capture the spatial organization of chromatin. Normalization of Hi-C contact map is essential for accurate modeling and interpretation of high-throughput chromatin conformation capture (3C) experiments. Hi-C correction tools were originally developed to normalize systematic biases of karyotypically normal cell lines. However, a vast majority of available Hi-C datasets are derived from cancer cell lines that carry multi-level DNA copy number variations (CNVs). CNV regions display over- or under-representation of interaction frequencies compared to CN-neutral regions. Therefore, it is necessary to remove CNV-driven bias from chromatin interaction data of cancer cell lines to generate a euploid-equivalent contact map. Results: We developed the HiCNAtra framework to compute high-resolution CNV profiles from Hi-C or 3C-seq data of cancer cell lines and to correct chromatin contact maps from systematic biases including CNV-associated bias. First, we introduce a novel ‘entire-fragment’ counting method for better estimation of the read depth (RD) signal from Hi-C reads that recapitulates the whole-genome sequencing (WGS)-derived coverage signal. Second, HiCNAtra employs a multimodal-based hierarchical CNV calling approach, which outperformed OneD and HiNT tools, to accurately identify CNVs of cancer cell lines. Third, incorporating CNV information with other systematic biases, HiCNAtra simultaneously estimates the contribution of each bias and explicitly corrects the interaction matrix using Poisson regression. HiCNAtra normalization abolishes CNV-induced artifacts from the contact map generating a heatmap with homogeneous signal. When benchmarked against OneD, CAIC, and ICE methods using MCF7 cancer cell line, HiCNAtra-corrected heatmap achieves the least 1D signal variation without deforming the inherent chromatin interaction signal. Additionally, HiCNAtra-corrected contact frequencies have minimum correlations with each of the systematic bias sources compared to OneD’s explicit method. Visual inspection of CNV profiles and contact maps of cancer cell lines reveals that HiCNAtra is the most robust Hi-C correction tool for ameliorating CNV-induced bias. Conclusions: HiCNAtra is a Hi-C-based computational tool that provides an analytical and visualization framework for DNA copy number profiling and chromatin contact map correction of karyotypically abnormal cell lines. HiCNAtra is an open-source software implemented in MATLAB and is available at https://github.com/AISKhalil/HiCNAtra. Ministry of Education (MOE) Nanyang Technological University Published version This work was supported by the Nanyang Technological University’s Nanyang Assistant Professorship grant and Singapore Ministry of Education Academic Research Fund Tier 1 grants (RG46/16 and RG39/18) to AS. AC is supported by the Nanyang Technological University start-up grant. The funding bodies were not involved in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.

Sensorimotor control studies have predominantly focused on how motor regions of the brain relay basic movement-related information such as position and velocity. However, motor control is often complex, involving the integration of sensory information, planning, visuomotor tracking, spatial mapping, retrieval and storage of memories, and may even be emotionally driven. This suggests that many more regions in the brain are involved beyond premotor and motor cortices. In this study, we exploited an experimental setup wherein activity from over 87 non-motor structures of the brain were recorded in eight human subjects executing a center-out motor task. The subjects were implanted with depth electrodes for clinical purposes. Using training data, we constructed subject-specific models that related spectral power of neural activity in six different frequency bands as well as a combined model containing the aggregation of multiple frequency bands to movement speed. We then tested the models by evaluating their ability to decode movement speed from neural activity in the test data set. The best models achieved a correlation of 0.38 ± 0.03 (mean ± standard deviation). Further, the decoded speeds matched the categorical representation of the test trials as correct or incorrect with an accuracy of 70 ± 2.75% across subjects. These models included features from regions such as the right hippocampus, left and right middle temporal gyrus, intraparietal sulcus, and left fusiform gyrus across multiple frequency bands. Perhaps more interestingly, we observed that the non-dominant hemisphere (ipsilateral to dominant hand) was most influential in decoding movement speed.

Winter dens are one of the important components of brown bear's (Ursus arctos syriacus) habitat, affecting their reproduction and survival. Therefore identification of factors affecting the habitat selection and suitable denning areas in the conservation of our largest carnivore is necessary. We used Geographically Weighted Logistic Regression (GWLR) and Generalized Linear Model (GLM) for modeling suitability of denning habitat in Kouhkhom region in Fars province. In the present research, 20 dens (presence locations) and 20 caves where signs of bear were not found (absence locations) were used as dependent variables and six environmental factors were used for each location as independent variables. The results of GLM showed that variables of distance to settlements, altitude, and distance to water were the most important parameters affecting suitability of the brown bear's denning habitat. The results of GWLR showed the significant local variations in the relationship between occurrence of brown bear dens and the variable of distance to settlements. Based on the results of both models, suitable habitats for denning of the species are impassable areas in the mountains and inaccessible for humans.

Precise diagnosis of psychiatric diseases and a comprehensive assessment of a patient's symptom severity are important in order to establish a successful treatment strategy for each patient. Although great efforts have been devoted to searching for diagnostic biomarkers of schizophrenia over the past several decades, no study has yet investigated how accurately these biomarkers are able to estimate an individual patient's symptom severity. In this study, we applied electrophysiological biomarkers obtained from electroencephalography (EEG) analyses to an estimation of symptom severity scores of patients with schizophrenia. EEG signals were recorded from 23 patients while they performed a facial affect discrimination task. Based on the source current density analysis results, we extracted voxels that showed a strong correlation between source activity and symptom scores. We then built a prediction model to estimate the symptom severity scores of each patient using the source activations of the selected voxels. The symptom scores of the Positive and Negative Syndrome Scale (PANSS) were estimated using the linear prediction model. The results of leave-one-out cross validation (LOOCV) showed that the mean errors of the estimated symptom scores were 3.34 ± 2.40 and 3.90 ± 3.01 for the Positive and Negative PANSS scores, respectively. The current pilot study is the first attempt to estimate symptom severity scores in schizophrenia using quantitative EEG features. It is expected that the present method can be extended to other cognitive paradigms or other psychological illnesses.

There is today a broad consensus in the scientific community regarding the involvement of acoustic cues in speech perception. Up to now, however, the precise mechanisms underlying the transformation from continuous acoustic stream into discrete linguistic units remain largely undetermined. This is partly due to the lack of an effective method for identifying and characterizing the auditory primitives of speech. Since the earliest studies on the acoustic–phonetic interface by the Haskins Laboratories in the 50’s, a number of approaches have been proposed; they are nevertheless inherently limited by the non-naturalness of the stimuli used, the constraints of the experimental apparatus, and the a priori knowledge needed. The present thesis aimed at introducing a new method capitalizing on the speech-in-noise situation for revealing the acoustic cues used by the listeners.As a first step, we adapted the Classification Image technique, developed in the visual domain, to a phoneme categorization task in noise. The technique relies on a Generalized Linear Model to link each participant’s response to the specific configuration of noise, on a trial-by-trail basis, thereby estimating the perceptual weighting of the different time-frequency regions for the decision. We illustrated the effectiveness of our Auditory Classification Image method through 2 examples: a /aba/-/ada/ categorization and a /da/-/ga/ categorization in context /al/ or /aʁ/. Our analysis confirmed that the F2 and F3 onsets were crucial for the tasks, as suggested in previous studies, but also revealed unexpected cues. In a second step, we relied on this new method to compare the results of musical experts (N=19) or dyslexics participants (N=18) to those of controls. This enabled us to explore the specificities of each group’s listening strategies.All the results taken together show that the Auditory Classification Image method may be a more precise and more straightforward approach to investigate the mechanisms at work at the acoustic-phonetic interface.; Bien qu’il existe un large consensus de la communauté scientifique quant au rôle des indices acoustiques dans la compréhension de la parole, les mécanismes exacts permettant la transformation d’un flux acoustique continu en unités linguistiques élémentaires demeurent aujourd’hui largement méconnus. Ceci est en partie dû à l’absence d’une méthodologie efficace pour l’identification et la caractérisation des primitives auditives de la parole. Depuis les premières études de l’interface acoustico-phonétique par les Haskins Laboratories dans les années 50, différentes approches ont été proposées ; cependant, toutes sont fondamentalement limitées par l’artificialité des stimuli utilisés, les contraintes du protocole expérimental et le poids des connaissances a priori nécessaires. Le présent travail de thèse s’est intéressé { la mise en oeuvre d’une nouvelle méthode tirant parti de la situation de compréhension de parole dégradée pour mettre en évidence les indices acoustiques utilisés par l’auditeur.Dans un premier temps, nous nous sommes appuyés sur la littérature dans le domaine visuel en adaptant la méthode des Images de Classification à une tâche auditive de catégorisation de phonèmes dans le bruit. En reliant la réponse de l’auditeur { chaque essai à la configuration précise du bruit lors de cet essai, au moyen d’un Modèle Linéaire Généralisé, il est possible d’estimer le poids des différentes régions temps-fréquence dans la décision. Nous avons illustré l’efficacité de notre méthode, appelée Image de Classification Auditive, à travers deux exemples : une catégorisation /aba/-/ada/, et une catégorisation /da/-/ga/ en contexte /al/ ou /aʁ/. Notre analyse a confirmé l’implication des attaques des formants F2 et F3, déjà suggérée par de précédentes études, mais a également permis de révéler des indices inattendus. Dans un second temps, nous avons employé cette technique pour comparer les résultats de participants musiciens experts (N=19) ou dyslexiques (N=18) avec ceux de participants contrôles. Ceci nous a permis d’étudier les spécificités des stratégies d’écoute de ces différents groupes.L’ensemble des résultats suggèrent que les Images de Classification Auditives pourraient constituer une nouvelle approche, plus précise et plus naturelle, pour explorer et décrire les mécanismes { l’oeuvre au niveau de l’interface acoustico-phonétique.

Background In response to an unexpected long cold spell in February 1996 which killed more than 100 older adults (mostly living alone) in Hong Kong, the Hong Kong Senior Citizen Home Safety Association established a Personal Emergency Link Service to provide emergency contact to the older adults, which uses a telephone system to render emergency relief and total care service around the clock. To facilitate the dynamic and efficient allocation of service resources, it is crucial to understand the factors linked with use of the services and number of hospital admissions arising from PE link service. Methods We initially use the Poisson generalized linear model (GLM) with polynomial effect functions of relevant covariates. If the time series of residuals from fitting the Poisson GLM reveals significant serial correlation, a Poisson generalized linear autoregressive moving average (GLARMA) model is refitted to the data to account for the auto-correlation among the time series of daily call numbers. If the data is overdispersed relative to the best fitting Poisson GLARMA model, then the negative binomial GLARMA model is refitted to account for any overdispersion. In all the models, dummy variables for weekdays and months are included to account for any cyclic trends due weekday effect or month of the year effect. The secular time trend is modeled by a polynomial function of calendar time over the study period. Finally any critical temperatures are identified by visually inspecting the graph of the effect function of temperature. Results The weekday and month effects are both significant with Monday seeing more PE Link calls than Sunday and June seeing less than January. Temperature has significant effect on the PE Link call rate with the effect highly nonlinear. A critical temperature, below which excessive increase in PE link calls that lead to hospital admissions, is identified to be around 15 °C. Conclusion Identifying a threshold temperature which generates an excessive increase in the expected number of PE Link calls would be useful in service provision planning and support for elderly in need of hospital admission.

The main objective of this paper is to provide a new methodology for road safety assessment in Oman through the development of suitable accident prediction models. GLM technique with Poisson or NBR using SAS package was carried out to develop these models. The paper utilized the accidents data of 31 un-signalized T-intersections during three years. Five goodness-of-fit measures were used to assess the overall quality of the developed models. Two types of models were developed separately; the flow-based models including only traffic exposure functions, and the full models containing both exposure functions and other significant geometry and traffic variables. The results show that, traffic exposure functions produced much better fit to the accident data. The most effective geometric variables were major-road mean speed, minor-road 85th percentile speed, major-road lane width, distance to the nearest junction, and right-turn curb radius. The developed models can be used for intersection treatment or upgrading and specify the appropriate design parameters of T-intersections. Finally, the models presented in this thesis reflect the intersection conditions in Oman and could represent the typical conditions in several countries in the middle east area, especially gulf countries., {"references":["Peden, M., Toryan, T., Harvey, A., Bartolomeos, Ki., Laych, K., et al.\n(2009). \"Global Status Report on Road Safety, Time For Action\". World\nHealth Organization. Geneva.","Margaret Chan. (2013). \"Global Status Report on Road Safety, Time For\nAction\". World Health Organization. Geneva.","Montella, A. (2010).\" Analysis of Crash Contributory Factors at Urban\nRoundabouts.\" 89th Annual Meeting of Transportation Research Board,\nWashington, D.C.","Scully, J., and Fildes, B. (2006). \"An Evaluation of Intersection\nCharacteristics Associated With Crashes at Intersections in Melbourne\nCBD.\" Monish University Accident Research Center, Victoria.","Harnen, S., Radin Umar, R.S., Wong, S.V., and Wan Hashim, W. I.\n(2003) \"Motorcycle Crash Prediction Model for Non-Signalized\nIntersections.\" International Association of traffic and Safety Science\n(IATSS) Research, 27(2), 58-65.","Tay, R. (2008). \"Review of the CRISP Publicity Campaigns.\" Final\nReport Submitted to Capital Region Intersection Safety Partnership,\n(Online).www.drivetolive.ca/Downloads/Review_Campaigns_2008.pdf,\nRetrieved on/6/2008.","Farag, S. Gaber. (2011). \"Accident prediction models for road\nintersection in sultanate of Oman\". Msc Thesis, Minoufiya University,\nEgypt.","Sawalha, Z.A. and Sayed, T. (2006). \"Traffic Accident Modeling: Some\nStatistical Issues.\" Canadian Journal of Civil Engineering, 33(9), 1115-\n1124.","Hauer, E., Ng, J.C.N., and Lovell, J., and Morris, C.N. (1988).\n\"Estimation of Safety at Signalized Intersections\". Transportation\nResearch Record, Transportation Research Board, National Research\nCouncil, Washington, D.C., 1185, 46-61.\n[10] Jovanis, P.P, and Chang, H.L. (1986). \"Modeling the Relationship of\nAccidents to Miles Traveled.\" Transportation Research Record,\nTransportation Research Board, National Research Council,\nWashington, D.C., 1068, 42-51. [11] Miaou, S., and Lum, H. (1993). \"Modeling Vehicle Accident and\nHighway Geometric Design Relationships.\" Accident Analysis and\nPrevention, 25(6), 689-709.\n[12] Salifu, M. (2004). \"Accident Prediction Models for Unsignalised Urban\nJunctions in Ghana.\" International Association of traffic and Safety\nScience (IATSS) Research, 28(1), 68-81.\n[13] Kirolos Haleem and Mohamed Abdel-Aty(2010). \"Examining traffic\ncrash injury severity at unsignalized intersections\". Journal of safety\nresearch (Impact Factor: 1.34). 08/2010; 41(4):347-57.\nDOI:10.1016/j.jsr.2010.04.006.\n[14] K. Nachimuthu and P. Partheeban (2013).\" Development of A Road\nAccident Prediction Model Based on System Dynamics Approach,\nIndian Journal of Applied Research, Volume: 3 | Issue: 9 | Sept 2013 |\nISSN - 2249-555X.\n[15] Mohamed Abdel-Aty and Kirolos Haleem (2010). \"Analyzing angle\ncrashes at unsignalized intersections using machine learning techniques\"\nAccident; analysis and prevention (Impact Factor: 1.65). 01/2011;\n43(1):461-70. DOI:10.1016/j.aap.2010.10.002\n[16] Dhofar Municipality. (2006). \"Salalah Road Network Traffic Movement\nStudy.\" Office of the Minister of State and Governor of Dhofar.\n[17] Kulmala, R. And Roine, M. (1988). \"Accident Prediction Models for\nTwo-Lane Roads in Finland.\" Proceedings of the Conference on Traffic\nSafety Theory and Research Methods, Session 4: Statistical Analysis\nand Models. Amsterdam: SWOV, 89-103.\n[18] Kulmala, R. (1995). \"Safety at Rural Three and Four-Arm Junctions:\nDevelopment of Accident Prediction Models.\" Technical Research\nCentre of Finland, VTT Publications 233.\n[19] Abdel-Aty, M., and Radwan, A. E. (2000). \"Modeling Traffic Accident\nOccurrence and Involvement.\" Accident Analysis and Prevention, 32(5),\n633-642.\n[20] SAS. (2003). \"SAS Online 9.1.\" SAS Institute INC. Cary, NC, USA.\n[21] Agresti, A. (1996). \"An Introduction to Categorical Data Analysis.\" John\nWilly, New York.\n[22] Fridstorm, L., Ifver, J., Ingebrigesten, S., Kulmala, R., and Thomeson,\nL. K. (1995). \"Measuring the Contribution of Randomness, Exposure,\nWeather, and Day Light to the Variation in Road Accidents Counts.\"\nAccidents Analysis and Prevention, 27(1), 1-20.\n[23] Ben-Akiva, M., and Lerman, S. R (1985). Discrete Chice Analysis:\nTheory and Application to Travel Demond, the MIT Press,\nMassachusetts."]}

Máster en Matemáticas y Computación

In this paper, the Bayesian framework is used for the analysis of functional MRI (fMRI) data. Two algorithms are proposed to deal with the nonstationarity of the noise. The first algorithm is based on the temporal analysis of the data, while the second algorithm is based on the spatiotemporal analysis. Both algorithms estimate the variance of the noise across the images and the voxels. The first algorithm is based on the generalized linear model (GLM), while the second algorithm is based on a spatiotemporal version of it. In the GLM, an extended design matrix is used to deal with the presence of the drift in the fMRI time series. To estimate the regression parameters of the GLM as well as the variance components of the noise, the variational Bayesian (VB) methodology is employed. The use of the VB methodology results in an iterative algorithm, where the estimation of the regression coefficients and the estimation of variance components of the noise, across images and voxels, are interchanged in an elegant and fully automated way. The performance of the proposed algorithms (under the assumption of different noise models) is compared with the weighted least-squares (WLSs) method. Results using simulated and real data indicate the superiority of the proposed approach compared to the WLS method, thus taking into account the complex noise structure of the fMRI time series. Ieee Transactions on Information Technology in Biomedicine

The mobile phone network has been rapidly and widely growing and becoming one of the major communication channels for the day-to-day life in Thailand. Not only the rapidly development and improvement but also high competition in the mobile phone industry has resulted in the three major players in the Thai market, including AIS, DTAC, and TRUE service provider. This study emphasizes on factors influencing decision to subscribe mobile phone service and attitude toward mobile phone concerning on identity, usage, and expenses aspects. A sample of 600 respondents is observed and analysed using multivariate statistical analysis. Analyses of factors determining attitude toward mobile phone employ linear regression model and Generalized Linear Model (GLM) assuming Logit. The models applied in analysing decision to subscribe mobile phone service consist of 3 separate Probit models and Multivariate Probit for the three service providers. The results indicate that mobile phone network subscribers make decision to subscribe mobile phone service by taking all three services providers as choices simultaneous, thus, Multivariate Probit models show significant estimated results. The findings also reveal that attitude towards mobile phone and marketing mixed have significant impacts on decision to subscribe mobile phone service.

The study emphasizes on factors determining stakeholders’ satisfaction towards financial disclosure of the listed companies in Stock Exchange of Thailand (SET). The factors include personal characteristics, media exposure, problem experienced from using financial information, investment behaviour, expectation, and attitude. A sample of 276 respondents representing stakeholders of the listed companies in SET is observed and analyzed by employing Linear Regression Model, Generalized Linear Models, and Ordered Logit Models. The results from Linear Regression Models and Ordered Logit Models indicated that level of media exposure, problem experienced, expectation and attitude have significant impacts on level of satisfaction towards financial disclosure of the listed companies in Stock Exchange of Thailand. The results also reveal that investment strategy of the respondents has significant impact on level of satisfaction. Respondents with fundamental based investment strategy are more likely to have lower level of satisfaction compared to technical based strategy.

Increasing in oil prices and critical condition in climate changes have put the pressure on the evolution of energy saving automobiles. Thai government has launched the policy to induce car users to switch to use eco-cars. However, according to the energy saving condition, capability of eco-car is limited to some certain level. With this limitation, car users might perceive eco-cars as inferior. This study emphasizes on car users’ attitude toward ecocars and factors that influence their attitude as well as the chance of using eco-cars. Sample of 560 respondents of car users in Bangkok, Thailand is collected and analyzed using multivariate analysis. Linear regression and Generalized Linear Model assuming Logit are employed in analyzing factor determining attitude toward eco-cars. Linear regression model and Ordered Logit model are applied in determining factor influencing chance of using eco-cars. The results confirm the hypothesis that attitude of car users on eco-cars plays an important role in determine decision of using eco-cars. Thus, Thai government should not only promote eco-cars through the low price and energy saving incentive policies but also the policy to change the car users’ attitude toward eco-cars.

We develop a generalized linear mixed model (GLMM) for bivariate count responses for statistically analyzing dragonfly population data from the Northern Netherlands. The populations of the threatened dragonfly species Aeshna viridis were counted in the years 2015–2018 at 17 different locations (ponds and ditches). Two different widely applied population size measures were used to quantify the population sizes, namely the number of found exoskeletons (‘exuviae’) and the number of spotted egg-laying females were counted. Since both measures (responses) led to many zero counts but also feature very large counts, our GLMM model builds on a zero-inflated bivariate geometric (ZIBGe) distribution, for which we show that it can be easily parameterized in terms of a correlation parameter and its two marginal medians. We model the medians with linear combinations of fixed (environmental covariates) and random (location-specific intercepts) effects. Modeling the medians yields a decreased sensitivity to overly large counts; in particular, in light of growing marginal zero inflation rates. Because of the relatively small sample size (n = 114) we follow a Bayesian modeling approach and use Metropolis-Hastings Markov Chain Monte Carlo (MCMC) simulations for generating posterior samples.