Search

Super sweet corn breeding must develop hybrids that fully meet the expectations of the market. In this sense, the determination of keytraits for the selection of sweetcorn genotypes is a fundamental condition for breeding success. The objective of this work was to identify key traits for the selection of promising and contrasting genotypes of super sweet corn. The experiments were carried out in Guarapuava (PR), in two sowing dates. Seventeen traits related to the yield and quality of the ear were evaluated. To perform the multivariate analysis, data were subjected to diagnosis of multicolinearity, analysis of canonical variables, genetic divergence, hierarchical clusters, factor analysis, and canonical correspondence analysis. Grain yield, yield of dehusked ears, and number of commercial ears were considered related key traits in the identification of promising super sweet corn genotypes. Hybrid D3-20 × D5-41 presented higher averages than the others, considering the yield related keytraits.

Bu çalışmanın amacı regresyon modellerinde çoklu doğrusal bağlantı olması durumunda işlevini yitiren en küçük kareler (EKK) metodu ile yanlı tahmin ediciler olan Ridge Regresyon (RR) ve Temel Bileşenler Regresyonu (TBR) yöntemlerini karşılaştırmaktır. Bu amaçla Tokat Gaziosmanpaşa Üniversitesi Araştırma ve Uygulama Çiftliğinde yetiştirilen sütten kesim döneminde bulunan 59 baş Saanen oğlaklarının canlı ağırlıkları (CA) ve bazı vücut özelliklerinden vücut uzunluğu (VU), cidago yüksekliği (CY), sağrı yüksekliği (SY), göğüs derinliği (GD), göğüs genişliği (GG) ve göğüs çevresi (GÇ) ölçümleri kullanılmıştır. Metotların tahmin performansını değerlendirmek için belirtme katsayısı (R2 ) ve hata kareler ortalaması (HKO) değerleri incelenmiştir. Çalışmada, canlı ağırlık tahmininde kullanılan cidago yüksekliği (CY) ve sağrı yüksekliği (SY) arasında çoklu doğrusal bağlantı olduğu görülmüş ve bu sorun RR ve TBR ile ortadan kaldırılmıştır. İncelenen yöntemlerin R2 ve HKO değerleri karşılaştırıldığında; Saanen oğlaklarının canlı ağırlık tahmininde RR yaklaşımının diğer yöntemlerden daha iyi sonuç verdiği belirlenmiştir. The aim of this study is to compare the least squares (LS) method that lost its function in the case of multicollinearity in regression methods with Ridge Regression (RR) and Principal Components Regression (PCR) which are bias estimators. For this aim, the effect of some body measurements on body weight (BW), body length (BL), height at withers (HW), height at rump (HR), chest depth (CD), chest girth (CG) and chest width (CW) obtained from 59 Saanen kids at weaning period raised at Research Farm of Tokat Gaziosmanpaşa University. Determination coefficient (R2 ) and mean square error (MSE) values were used to evaluate the estimation performance of the methods. The multicollinearity between height at withers (HW) and height at rump (HR) which were used to estimate body weight was eliminated by using RR and PCR. When R2 and HKO values of the examined methods are compared; It has been shown that RR method have better results of live weight of Saanen goats.

The vessel neutron irradiation, which cannot be replaced, is one of the limiting factors for pressurized water reactor (PWR) lifetime. Surveillance programmes are therefore necessary for safety assessment and for verifying the vessel structural integrity. The quality of radiation damage prediction depends in part on the calculation of the fast neutron flux. In that sense, a lack of knowledge on the fast neutron flux will require larger safety margins on the plant lifetime affecting operating conditions and the cost of nuclear installations. To make correct decisions when designing the plant lifetime and on safety margins for PWR reactors, it is therefore essential to assess the uncertainty in vessel flux calculations. Most of the past studies on the flux uncertainty quantification are based on the methods of moments which assumes a linear output variation. This method was most commonly used because the calculation capabilities of computers prevented from conducting more accurate methods. In a non-linear case, the first order hypothesis appears insufficient for an accurate prediction of the output variance.An alternative method is the Total Monte Carlo approach (TMC) which consists in randomly sampling the input data and propagating the perturbations on the calculation chain. The advantage of this method is that it does not make any assumptions on the linear interactions or small input changes among data. It considers the probability distributions of input parameters and thus provides a more precise description of input uncertainties.It is within this context that this thesis was conducted. It consists in performing a new uncertainty assessment of the fast flux calculation for the PWR vessel considering the data of recent international nuclear libraries. The special feature of this thesis lies in the large number of uncertain parameters which are closely correlated with each other. The uncertainty on the fast flux, considering all the uncertain parameters, is finally estimated for the vessel hot spot. More generally, in this context of sensitivity analysis, we show the importance to consider the covariance matrices to propagate the input uncertainties, and to analyze the contribution of each input on a physical model. The Shapley and Johnson indices are especially used in a multicolinearity context between the inputs and the output., Pour s’assurer de la sûreté des réacteurs et de l’intégrité de leurs cuves, des programmes de surveillance sont mis en place. Ces derniers dépendent notamment du calcul de densité neutronique qui quantifie le nombre de neutrons rapides (susceptibles de modifier les propriétés mécaniques de la cuve) reçus par les capsules de surveillance. Or, un manque de connaissances sur ce calcul peut obliger l’exploitant à fixer des marges de sûreté supérieures sur la durée de vie des centrales, et donc affecter les conditions de fonctionnement et le coût des installations nucléaires. C’est pourquoi il est essentiel de déterminer l’incertitude du calcul de flux à la cuve des REP. La plupart des études passées sont basées sur une quantification de l’incertitude par la méthode des moments qui suppose une variation linéaire de la sortie du calcul. Cette hypothèse n’a jamais été vérifiée compte tenu du grand nombre de variables d’entrée et des ressources informatiques relativement limitées pour envisager l’application d’autres méthodes. Une méthode alternative est celle de l’approche Total Monte Carlo (TMC) qui consiste à considérer les paramètres d’entrées du modèle de calcul comme des variables aléatoires, et à propager les perturbations issues d’un tirage statistique dans la chaîne de calcul. L’avantage de cette méthode est qu’elle propose une description plus précise des incertitudes du modèle, et permet de définir des quantiles et donc des intervalles de confiance qui peuvent s’avérer utiles dans le cadre d'une étude de sûreté. C’est dans ce contexte, que cette thèse s’intègre. Elle consiste à déterminer l’incertitude du flux neutronique rapide de la cuve des REP en considérant les librairies de données nucléaires récentes et en utilisant une approche TMC pour la propagation des incertitudes d’entrées. La particularité de ce travail repose sur le grand nombre de paramètres d’entrées corrélés qui nécessite la mise en place de méthodologies adaptées. En particulier, les indices de Shapley et de Johnson qui sont peu connus dans le domaine du nucléaire, sont proposés dans le cadre de l'analyse globale de sensibilité d'un problème multicolinéaire.

Some procedures can be used for selecting independent variables, one of them is the procedure of all possible regression with robust Cp (RCp) criterion. This statistic is not sensitive with multicollinearity in model and outlier residuals. The aim of this article is to investigate the use of RCp criterion in selecting independent variables. The result of the simulation experimental data shows that the RCp criterion fits enough to select independent variables.

Aflatoxin is a carcinogenic toxin to humans and animals produced by mold fungi in staple crops. Surveys of Aflatoxin are expensive, and the results are usually not available for implementing within season mitigation strategies. Identification of high and low risk areas and years is essential to reduce the number of samples analyzed for Aflatoxin concentration. Previously a risk factors approach was developed to determine county level Aflatoxin contamination risk in southern Georgia, but Aflatoxin concentrations and risk factor data were not analyzed simultaneously and all risk factors had equal weight which is unrealistic. In the current paper we propose a regression approach to overcome these problems. Spatial Poisson profile regression identified clusters of counties which have similar Aflatoxin risk and risk factor profiles, whilst explicitly taking into account multicollinearity in the risk factor data and spatial autocorrelation in the Aflatoxin data. This approach allows examination of the utility of different highly correlated variables including remotely sensed data that could give information at the sub-county level. The results identify plausible clusters compared to previous work but also give the relative importance of the risk factors associated with those clusters. The approach also helps show that some factors like well-drained soil behave differently from expectations and irrigation data is not useful.

У статті проведено математико-статистичний аналіз бізнес-індикаторів українських компаній з ризикового страхування з метою сприяння вибору релевантних методів оцінки фінансового здоров’я українських страховиків. Зокрема досліджено описову статистику обраних показників, перевірено відповідність характеристик даних основним припущенням ймовірнісних статистичних дискримінаційних методів. На основі проведеного аналізу запропоновано перелік змінних та методів, що можуть бути обґрунтовано використані в оцінці фінансового стану українських страховиків. В статье проведен математико-статистический анализ бизнес-индикаторов украинских компаний по рисковому страхованию с целью способствования выбору релевантных методов оценки финансового здоровья украинских страховщиков. В частности исследована описательная статистика выбранных показателей, проверено соответствие характеристик данных основным предположениям вероятностных статистических дискриминационных методов. На основании проведенного анализа предложен перечень переменных и методов, которые могут быть обоснованно использованы в оценке финансового состояния украинских страховщиков. In the article, a mathematical-statistical analysis of Ukrainian general insurance companies’ business indicators has been performed in order to promote the selection of proper methods of evaluating the financial health of Ukrainian insurers. In particular, descriptive statistics have been investigated, and data characteristics checked against the requirements of probabilistic statistical discriminative methods. Based on the analysis, a list of indicators and methods that can be reasonably used in evaluation of financial condition of Ukrainian insurers have been proposed.

Списание "Бизнес посоки" 2013 г. брой 2, В статията се разглеждат три въпроса, свързани с приложението на регресионния анализ за измерване на ефектите от насърчаване на продажбите. Изборът на зависима променлива определя вида на ефектите, които могат да се измерят. Наличието на няколко независими променливи повдига въпроса за тяхното взаимодействие. Ефектите, свързани с промяна на времето на покупка предполагат включването на лагови променливи в регресионния модел. This article puts three issues related to the application of regression analysis to sales promotion effects measurement. The choice of the dependent variable determined the type of effects that can be measured. The presence of several independent variables raises questions about their interaction. Effects associated with purchase timing suggest the inclusion of lagged variables in the regression model.

Kianoush Fathi Vajargah,1 Homayoun Sadeghi-Bazargani,2,3 Robab Mehdizadeh-Esfanjani,4 Daryoush Savadi-Oskouei,4 Mehdi Farhoudi41Department of Statistics, Islamic Azad University, Tehran, North Branch, 2Neuroscience Research Center, Department of Statistics and Epidemiology, Tabriz University of Medical Sciences, Tabriz, Iran; 3Public Health Department, Karolinska Institute, Stockholm, Sweden; 4Neuroscience Research Center, Tabriz University of Medical Sciences, Tabriz, IranAbstract: The objective of the present study was to assess the comparable applicability of orthogonal projections to latent structures (OPLS) statistical model vs traditional linear regression in order to investigate the role of trans cranial doppler (TCD) sonography in predicting ischemic stroke prognosis. The study was conducted on 116 ischemic stroke patients admitted to a specialty neurology ward. The Unified Neurological Stroke Scale was used once for clinical evaluation on the first week of admission and again six months later. All data was primarily analyzed using simple linear regression and later considered for multivariate analysis using PLS/OPLS models through the SIMCA P+12 statistical software package. The linear regression analysis results used for the identification of TCD predictors of stroke prognosis were confirmed through the OPLS modeling technique. Moreover, in comparison to linear regression, the OPLS model appeared to have higher sensitivity in detecting the predictors of ischemic stroke prognosis and detected several more predictors. Applying the OPLS model made it possible to use both single TCD measures/indicators and arbitrarily dichotomized measures of TCD single vessel involvement as well as the overall TCD result. In conclusion, the authors recommend PLS/OPLS methods as complementary rather than alternative to the available classical regression models such as linear regression.Keywords: Prognostic study, trans cranial doppler, partial least squares regression, orthogonal projections to latent structures, multicolinearity

Many disclosure and determinants studies have been conducted, in which the authors most often estimate a classical linear regression model (CLR). This technique is the best technique in order to estimate linear relationships between a dependent variable and some independent variables, when the constraints that apply to such a model are respected. The lack of a multicolinearity problem being one of these constraints, the aim of this article is 1/ to show that the usual tools that are used in order to detect and remedy this problem are sometimes not sufficient 2/ to demonstrate the superiority of the indicators that were proposed by Besley, Kuh and Welsch (1980). From these indicators, an algorithm is proposed in this article which is then applied in an illustrative way in a determinant and disclosure study. We then show that this algorithm allows circumventing a multicolinearity problem., De nombreuses études ont été réalisées sur le thème des déterminants d'une publication volontaire d'informations et la plupart des auteurs de ces études recourent à l'estimation d'un modèle de régression linéaire classique (RLC). Cette technique aboutit en effet, lorsque les contraintes nécessaires à son application sont respectées, à l'utilisation des meilleurs estimateurs linéaires sans biais. Une absence de multicolinéarité constituant l'une de ces contraintes, l'objet de cet article est 1/de montrer l'insuffisance relative des outils de diagnostic généralement utilisés 2/ de démontrer les avantages et la supériorité des indicateurs de Belsley, Kuh et Welsch (1980). A partir de ces indicateurs, cet article propose un algorithme de décision simplifié qui est appliqué d'une façon illustrative au sein d'une étude empirique portant sur les déterminants d'une publication volontaire d'informations sur les activités de recherche et développement.

The bachelor thesis aims to create a comprehensible user's manual for econometric software GRETL in the Czech language. It begins with the introduction of software GRETL and then acquaints the reader with a range of possible uses of the software when working with data. Thesis briefly describes the main characteristics of the possible linear regression model, such as multicollinearity, autocorrelation and heteroscedasticity, including possible testing in the particular software and evaluation of specific tests. The key part deals with defining and specifying data for their subsequent analysis in the software, compilation of a model, estimates of parameters of an econometric model using the least squares method, the weighted least squares method and the generalized least squares method. Furthermore, the thesis deals with hypotheses testing and confidence interval estimations, and also shows how to create and edit graphs in GRETL and explains each menu item in detail. All steps are supported by a graphical supplement and specific examples.

Submitted by Guilherme Lemeszenski (guilherme@nead.unesp.br) on 2013-08-22T19:07:19Z No. of bitstreams: 1 S1516-35982008000800007.pdf: 62982 bytes, checksum: c2cb745b81a3f4ccfee59935136f45f7 (MD5) Made available in DSpace on 2013-08-22T19:07:19Z (GMT). No. of bitstreams: 1 S1516-35982008000800007.pdf: 62982 bytes, checksum: c2cb745b81a3f4ccfee59935136f45f7 (MD5) Previous issue date: 2008-08-01 Made available in DSpace on 2013-09-30T20:11:45Z (GMT). No. of bitstreams: 2 S1516-35982008000800007.pdf: 62982 bytes, checksum: c2cb745b81a3f4ccfee59935136f45f7 (MD5) S1516-35982008000800007.pdf.txt: 43114 bytes, checksum: 99ddd894c644222a4140b62043a7c9ff (MD5) Previous issue date: 2008-08-01 Submitted by Vitor Silverio Rodrigues (vitorsrodrigues@reitoria.unesp.br) on 2014-05-20T15:18:14Z No. of bitstreams: 2 S1516-35982008000800007.pdf: 62982 bytes, checksum: c2cb745b81a3f4ccfee59935136f45f7 (MD5) S1516-35982008000800007.pdf.txt: 43114 bytes, checksum: 99ddd894c644222a4140b62043a7c9ff (MD5) Made available in DSpace on 2014-05-20T15:18:14Z (GMT). No. of bitstreams: 2 S1516-35982008000800007.pdf: 62982 bytes, checksum: c2cb745b81a3f4ccfee59935136f45f7 (MD5) S1516-35982008000800007.pdf.txt: 43114 bytes, checksum: 99ddd894c644222a4140b62043a7c9ff (MD5) Previous issue date: 2008-08-01 Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) Objetivou-se avaliar os efeitos aditivos direto e materno e os efeitos não-aditivos e heterozigóticos sobre os desempenhos pré e pós-desmama e desenvolver equações para predizer o desempenho de genótipos de uma população multirracial Aberdeen Angus (Angus) × Nelore. Nas análises, foram utilizados dois modelos. O modelo 1 foi adotado para as características pré-desmama peso à desmama ajustado para 205 dias (P205) e ganho de peso médio diário do nascimento à desmama (GMDND), considerando fixos os efeitos de grupo de contemporâneos (GC) e de grupo genético (GG), além das co-variáveis idade do bezerro à desmama (IBD) e idade da vaca ao parto (IVP). Para as características pós-desmama peso ao sobreano ajustado para 550 dias de idade (P550) e ganho médio de peso da desmama ao sobreano (GMDDS), utilizou-se o mesmo modelo, apenas substituindo a co-variável IBD pela idade do bezerro ao sobreano (IBS). O modelo 2 (M2), de regressão múltipla, foi utilizado depois que os dados foram ajustados para GC, IVP e IBD (P205) ou IBS (P550) e contemplou os efeitos genéticos aditivos diretos e maternos de raça e os efeitos de heterozigoses direta e materna. Os resultados das análises comprovaram diferenças entre os cruzamentos, uma vez que os animais com elevada heterozigose materna apresentaram melhor desempenho para P205 e GMDND, tendência semelhante à observada para a heterozigose direta. A influência do efeito direto sobre as características P550 e GMDDS foi maior que a do efeito materno. Os efeitos aditivos diretos da raça Angus determinaram maior desempenho. Não houve diferença significativa entre os valores estimados pelo modelo 1 e os preditos pelo modelo 2 para P205, P550 e GMDDS. The objective of this study was to estimate direct and maternal additive genetic, non-additive and heterozygotic effects for pre and post-weaning traits and to develop equations to predict growth performance in a multi-breed Aberdeen Angus (Angus)-Nellore herd. The analyses were performed by SAS® using two different models. The model 1 (M1) was used for the pre-weaning traits, weaning weight adjusted to 205 days of age (P205) and average daily gain from birth to weaning (GMDND), and included the fixed effects of contemporary groups (GC), genetic groups (GG) and the covariates age of calf at weaning (IBD) and age of cow at calving (IVP). The same model except for replacing the covariate IBD by age of calf at yearling (IBS) was used for the post-weaning traits - yearling weight adjusted to 550 days of age (P550) and average daily gain from weaning to yearling (GMDDS). A multiple regression model (M2) was used after adjusting the data for GC, IVP and IBD (P205) and IBS (P550) effects and included the breed direct and maternal additive and the direct and maternal heterozygotic effects. Animals with higher maternal heterozygosis performed better for P205 and GMDND. A similar trend was observed for the direct heterozygosis. The direct genetic effect was larger than the maternal genetic effect for P550 and GMDDS. The additive direct effect determined better performances in the Angus breed. Differences between the estimated (M1) and predicted (M2) averages for P205, P550 and GMDDS were not significant. UFSM UFSM Departamento de Zootecnia CNP PIBIC CAPES UNESP FCAV UNESP FCAV

This work deals with factor analysis method to investigate relationships among several data sets in veterinary epidemiology. Considering the aims and the nature of data collected within this framework, the research work is focused on the multiblock setting for the purpose of exploring and modelling one data set from several other data sets. The key feature is to handle the sensitivity of these methods to multicolinearity problem among the explanatory variables. Existing methods are described and unified by considering the criteria to be optimised or by setting up a continuum approach. New formulations of alternative methods which are robust to the multicolinearity problem are discussed. These methods can be applied to data organised in two or several data sets. The interest of the general strategy of analysis, related to the methods and continuums, is illustrated on the basis of real case studies pertaining to veterinary epidemiology., Ce travail de recherche s'inscrit dans le cadre des méthodes factorielles qui permettent de décrire et prédire des données structurées en plusieurs tableaux. Les objectifs et la nature des données d'épidémiologie analytique dans le domaine vétérinaire ont amené à centrer le travail sur les méthodes de régression multibloc, qui orientent la description de plusieurs tableaux de variables vers l'explication d'un autre tableau. Un des principaux objectifs est de contribuer à la réflexion sur la sensibilité de ces méthodes à la multicolinéarité. Des méthodes statistiques existantes sont présentées et reliées dans un cadre unifié, relevant soit de critères à maximiser comparables, soit d'un continuum général les reliant. De nouvelles méthodes peu vulnérables à l'égard de la multicolinéarité, et s'appliquant au cas de données structurées en deux puis en (K+1) tableaux, sont proposées. L'intérêt de ces méthodes, ainsi que des continuums qui leur sont associés, est illustré sur la base d'études de cas réels en épidémiologie. Ce travail de recherche a permis d'appliquer les méthodes multiblocs au domaine de l'épidémiologie animale, dans lequel elles n'avaient pas encore été utilisées.

This work builds up a complete procedure of econometric analysis of direct and indirect effects essential for monitoring performance indicators in manufacture sphere. By principal component method all performance indicators are extracted in appropriate homogeneous subsets. Paper is focused on estimation of standardized coefficient for each extracted factor, comparing their relative influence as well as their weights specification. Analytical hierarchy process is used as the technique of measuring inconsistency of assigned factor weights.

This paper deals with overcoming the situation when the multicolinearity appears as dominant problem in serious statistical-econometric research. Result of using standard statistical-econometric methods (as stepwise technique) is excluding a numerous explanatory variables with significant influence on dependent variable. Even more, multicolinearity becomes a barrier for specification of any influence of "removed" variables. This is especially relevant in the cases of analyzing total effects of the entire set of explanatory variables on the dependent variable. Moreover, indirect effects must not be ignored in the situation when they are dominant component of total effect. The whole procedure of relaxing this problem is illustrated by a practical example of comparing performance indicators of all manufacturing enterprises in the Split-Dalmatian County in 2004. The data base consists of a wide range of performance indicators for 1744 manufacturing enterprises, among which twelve are selected as representative ones. Using principal components method four factors have been extracted, i.e. all selected variables (performance indicators) have been meaningfully grouped in to factors: activity, liquidity, leverage, economic efficiency. It is shown that principal component method doesn’ t stands for a transformation method only, but it can be used as explicit modelling approach. These uncorrelated common factors are used to overcome multicolinearity by orthogonal transformation of explanatory variables in multiple regression model. The essential part of analysis is establishing of direct, indirect and overall effects of each explanatory variable on return on equity as chosen dependent variable. This work builds up a complete procedure of standardized coefficient estimation for each extracted factor, comparing their relative influence as well as their weights specification. Analytical hierarchy process is used as the technique of measuring inconsistency of assigned factor weights.

This work builds up a complete procedure of using factor analysis as the instrument in the case of relaxing the assumption of the classical model. Paper is focused on the situation when the multicolinearity appears as the dominant problem. This problem is solved by grouping of performance indicators, not only by technical principles, but also according to fundamental postulates of business economic theory. The whole procedure is illustrated by a practical example. The example originates from the real need to analyse and compare the performance of all manufacturing enterprises in the Split-Dalmatian County in 2004. The data set consists of a wide range of performance indicators for 1744 manufacturing enterprises, among which twelve are selected as representative ones. As the entire basic set is the issue of our interest, we find that the enterprises are markedly heterogeneous in terms of the chosen indicators. Therefore previous to comparison they have to be made homogeneous. After such homogenization, using principal components method four factors have been extracted, i.e. all selected variables (performance indicators) have been meaningfully grouped in to factors: activity, liquidity, leverage, economic efficiency. The essential part of analysis is establishing of direct, indirect and overall effects of each independent variable on return on equity as chosen dependent variable.