Dipòsit Digital de Documents de la UAB: sort
http://ddd.uab.cat
Dipòsit Digital de Documents de la UAB latest documents in sortcaWed, 23 Jul 2014 02:54:08 GMTInvenio 1.1.1ddd.bib@uab.cat360134125http://ddd.uab.cat/img/site_logo_rss.pngDipòsit Digital de Documents de la UAB
http://ddd.uab.cat
Search Search this site:p
http://ddd.uab.cat/search
Testing extreme value copulas to estimate the quantile
http://ddd.uab.cat/record/118912
We generalize the test proposed by Kojadinovic, Segers and Yan which is used for testing whether the data belongs to the family of extreme value copulas. We prove that the generalized test can be applied whatever the alternative hypothesis. We also study the effect of using different extreme value copulas in the context of risk estimation. To measure the risk we use a quantile. Our results have been motivated by a bivariate sample of losses from a real database of auto insurance claims. Bahraoui, ZuhairMon, 30 Jun 2014 12:58:59 GMThttp://ddd.uab.cat/record/118912Decision making in kidney paired donation programs with altruistic donors
http://ddd.uab.cat/record/118909
In recent years, kidney paired donation has been extended to include living non-directed or altruistic donors, in which an altruistic donor donates to the candidate of an incompatible donorcandidate pair with the understanding that the donor in that pair will further donate to the candidate of a second pair, and so on; such a process continues and thus forms an altruistic donor-initiated chain. In this paper, we propose a novel strategy to sequentially allocate the altruistic donor (or bridge donor) so as to maximize the expected utility; analogous to the way a computer plays chess, the idea is to evaluate different allocations for each altruistic donor (or bridge donor) by looking several moves ahead in a derived look-ahead search tree. Simulation studies are provided to illustrate and evaluate our proposed method. Li, YijiangMon, 30 Jun 2014 12:46:10 GMThttp://ddd.uab.cat/record/118909Regression analysis using order statistics and their concomitants
http://ddd.uab.cat/record/118908
In this work we derive the exact joint distribution of linear combinations of order statistics and linear combinations of their concomitants and some auxiliary variables in multivariate normal distribution. By extending the results of Sheikhi and Jamalizadeh we investigate some regression equations. Our results generalize those obtained in previous research by Viana, Lee and Loper-fido. Rasoul Ziaei, AbdulMon, 30 Jun 2014 12:37:13 GMThttp://ddd.uab.cat/record/118908Time-varying market beta :: does the estimation methodology matter?
http://ddd.uab.cat/record/118907
This paper compares the performance of nine time-varying beta estimates taken from three different methodologies never previously compared: least-square estimators including nonparametric weights, GARCH-based estimators and Kalman filter estimators. The analysis is applied to the Mexican stock market (2003-2009) because of the high dispersion in betas. The comparison between estimators relies on their financial applications: asset pricing and portfolio management. Results show that Kalman filter estimators with random coefficients outperform the others in capturing both the time series of market risk and their cross-sectional relation with mean returns, while more volatile estimators are better for diversification purposes. Nieto, BelénMon, 30 Jun 2014 12:30:01 GMThttp://ddd.uab.cat/record/118907Improving parametric Clarke and Wright algorithms by means of iterative empirically adjusted greedy heuristics
http://ddd.uab.cat/record/118906
Since Clarke and Wright proposed their well-known savings algorithm for solving the Capacitated Vehicle Routing Problem, several enhancements to the original savings formula have been recently proposed, in the form of parameterisations. In this paper we first propose to use Empirically Adjusted Greedy Heuristics to run these parameterized heuristics and we also consider the addition of new parameters. This approach is shown to improve the savings algorithms proposed in the literature. Moreover, we propose a new procedure which leads to even better solutions, based on what we call Iterative Empirically Adjusted Greedy Heuristics. Corominas, AlbertMon, 30 Jun 2014 12:22:02 GMThttp://ddd.uab.cat/record/118906Editor’s Report 2014
http://ddd.uab.cat/record/118905
Guillén Estany, MontserratMon, 30 Jun 2014 12:06:43 GMThttp://ddd.uab.cat/record/118905Analysis of inequality in fertility curves fitted by gamma distributions
http://ddd.uab.cat/record/117895
The aim of this paper is to analyse fertility curves from a novel viewpoint, that of inequality. Through sufficient conditions that can be easily verified, we compare inequality, in the Lorenz and Generalized Lorenz sense, in fertility curves fitted by gamma distributions, thus achieving a useful complementary instrument for demographic analysis. As a practical application, we examine inequality behaviour in the distributions of specific fertility curves in Spain from 1975 to 2009. Ramos, Héctor M.Mon, 05 May 2014 13:29:26 GMThttp://ddd.uab.cat/record/117895Double bounded Kumaraswamy-power series class of distributions
http://ddd.uab.cat/record/117894
In this paper, we will introduce the new Kumaraswamy-power series class of distributions. This new class is obtained by compounding the Kumaraswamy distribution of Kumaraswamy (1980) and the family of power series distributions. The new class contains some new double bounded distributions such as the Kumaraswamy-geometric, -Poisson, -logarithmic and -binomial, which are used widely in hydrology and related areas. In addition, the corresponding hazard rate function of the new class can be increasing, decreasing, bathtub and upside-down bathtub. Some basic properties of this class of distributions such as the moment generating function, moments and order statistics are studied. Some special members of the class are also investigated in detail. The maximum likelihood method is used for estimating the unknown parameters of the members of the new class. Finally, an application of the proposed class is illustrated using a real data set. Bidram, HamidMon, 05 May 2014 13:25:14 GMThttp://ddd.uab.cat/record/117894An alternative to Kim and Warde’s mixed randomized response model
http://ddd.uab.cat/record/117893
When open or direct surveys are about sensitive matters (e. g. gambling habits, addiction to drug and others intoxicants, alcoholism, proneness to tax invasion, induced abortions, drunken driving, history of past involvement in crimes, and homosexuality), non-response bias and response bias become serious problems because people oftentimes do not wish to give correct information. To reduce non-response and response bias, various alternative approaches have been proposed, for example a randomized response survey technique, or a mixed randomized response model using simple random sampling with a replacement sampling scheme that improves the privacy of respondents, proposed by authors Kim and Warde. In this paper we have suggested an alternative to Kim and Warde’s mixed randomized response model to estimate the proportion of qualitative sensitive variable under the conditions presented in both the cases of completely truthful reporting and less than completely truthful reporting by the respondents. Properties of the proposed randomized response model have been studied along with recommendations. We have also extended the proposed model to stratified random sampling. Numerical illustrations and graphs are also given in support of the present study. Singh, Housila P.Mon, 05 May 2014 13:21:40 GMThttp://ddd.uab.cat/record/117893Quantile estimation of the rejection distribution of food products integrating assessor values and interval-censored consumer data
http://ddd.uab.cat/record/117891
Fitting parametric survival models with interval-censored data is a common task in survival analysis and implemented in many statistical software packages. Here, we present a novel approach to fit such models if the values on the scale of interest are measured with error. Random effects ANOVA models are used to account for the measurement errors and the likelihood function of the parametric survival model is maximized with numerical methods. An illustration is provided with a real data set on the rejection of yogurt as a function of its acid taste. Langohr, KlausMon, 05 May 2014 13:17:54 GMThttp://ddd.uab.cat/record/117891A nonparametric visual test of mixed hazard models
http://ddd.uab.cat/record/117890
We consider mixed hazard models and introduce a new visual inspection technique capable of detecting the credibility of our model assumptions. Our technique is based on a transformed data approach, where the density of the transformed data should be close to the uniform distribution when our model assumptions are correct. To estimate the density on the transformed axis we take advantage of a recently defined local linear density estimator based on filtered data. We apply the method to national mortality data and show that it is capable of detecting signs of heterogeneity even in small data sets with substantial variability in observed death rates. Spreeuw, JaapMon, 05 May 2014 13:09:18 GMThttp://ddd.uab.cat/record/117890Selection and pattern mixture models for modelling longitudinal data with dropout: An application study
http://ddd.uab.cat/record/117889
Incomplete data are unavoidable in studies that involve data measured or observed longitudinally on individuals, regardless of how well they are designed. Dropout can potentially cause serious bias problems in the analysis of longitudinal data. In the presence of dropout, an appropriate strategy for analyzing such data would require the definition of a joint model for dropout and measurement processes. This paper is primarily concerned with selection and pattern mixture models as modelling frameworks that could be used for sensitivity analysis to jointly model the distribution for the dropout process and the longitudinal measurement process. We demonstrate the application of these models for handling dropout in longitudinal data where the dependent variable is missing across time. We restrict attention to the situation in which outcomes are continuous. The primary objectives are to investigate the potential influence that dropout might have or exert on the dependent measurement process based on the considered data as well as to deal with incomplete sequences. We apply the methods to a data set arising from a serum cholesterol study. The results obtained from these methods are then compared to help gain additional insight into the serum cholesterol data and assess sensitivity of the assumptions made. Results showed that additional confidence in the findings was gained as both models led to similar results when assessing significant effects, such as marginal treatment effects. Satty, AliMon, 05 May 2014 13:04:23 GMThttp://ddd.uab.cat/record/117889Locally adaptive density estimation on Riemannian manifolds
http://ddd.uab.cat/record/117888
In this paper, we consider kernel type estimator with variable bandwidth when the random variables belong in a Riemannian manifolds. We study asymptotic properties such as the consistency and the asymptotic distribution. A simulation study is also considered to evaluate the performance of the proposal. Finally, to illustrate the potential applications of the proposed estimator, we analyse two real examples where two different manifolds are considered. Henry, GuillermoMon, 05 May 2014 12:59:26 GMThttp://ddd.uab.cat/record/117888Flexible quantile regression models : application to the study of the purple sea urchin
http://ddd.uab.cat/record/113246
In many applications, it is often of interest to assess the possible relationships between covariates and quantiles of a response variable through a regression model. In some instances, the effects of continuous covariates on the outcome are highly nonlinear. Consequently, appropriate modelling. ; has to take such flexible smooth effects into account. In this work, various flexible quantile regression techniques were reviewed and compared by simulation. Finally, all the techniques were used to construct the overall zone specific reference curves of morphologic measures of sea urchin Paracentrotus lividus (Lamarck, 1816) located in NW Spain. Martínez-Silva, IsabelMon, 11 Nov 2013 14:57:28 GMThttp://ddd.uab.cat/record/113246New insights into evaluation of regression models through a decomposition of the prediction errors : application to near-infrared spectral data
http://ddd.uab.cat/record/113245
This paper analyzes the performance of linear regression models taking into account usual criteria such as the number of principal components or latent factors, the goodness of fit or the predictive capability. Other comparison criteria, more common in an economic context, are also considered:. ; the degree of multicollinearity and a decomposition of the mean squared error of the prediction which determines the nature, systematic or random, of the prediction errors. The applications use real data of extra-virgin oil obtained by near-infrared spectroscopy. The high dimensionality of the data is reduced by applying principal component analysis and partial least squares analysis. A possible improvement of these methods by using cluster analysis or the information of the relative maxima of the spectrum is investigated. Finally, obtained results are generalized via cross- validation and bootstrapping. Sánchez-Rodríguez, María IsabelMon, 11 Nov 2013 14:53:37 GMThttp://ddd.uab.cat/record/113245The normal distribution in some constrained sample spaces
http://ddd.uab.cat/record/113243
Phenomena with a constrained sample space appear frequently in practice. This is the case, for example, with strictly positive data, or with compositional data, such as percentages or proportions. If the natural measure of difference is not the absolute one, simple algebraic properties show that it is more convenient to work with a geometry different from the usual Euclidean geometry in real space, and with a measure different from the usual Lebesgue measure, leading to alternative models that better fit the phenomenon under study. The general approach is presented and illustrated using the normal distribution, both on the positive real line and on the D-part simplex. The original ideas of McAlister in his introduction to the lognormal distribution in 1879, are recovered and updated. Mateu-Figueras, GlòriaMon, 11 Nov 2013 14:39:59 GMThttp://ddd.uab.cat/record/113243A note on the Fisher information matrix for the skew-generalized-normal model
http://ddd.uab.cat/record/113242
In this paper, the exact form of the Fisher information matrix for the skew-generalized normal (SGN) distribution is determined. The existence of singularity problems of this matrix for the skewnormal and normal particular cases is investigated. Special attention is given to the asymptotic properties of the MLEs under the skew-normality hypothesis. Arellano-Valle, Reinaldo B.Mon, 11 Nov 2013 14:32:01 GMThttp://ddd.uab.cat/record/113242Improved entropy based test of uniformity using ranked set samples
http://ddd.uab.cat/record/113241
Ranked set sampling (RSS) is known to be superior to the traditional simple random sampling (SRS) in the sense that it often leads to more efficient inference procedures. Basic version of RSS has been extensively modified to come up with schemes resulting in more accurate estimators of the population attributes. Multistage ranked set sampling (MSRSS) is such a variation surpassing RSS. Entropy has been instrumental in constructing criteria for fitting of parametric models to the data. The goal of this article is to develop tests of uniformity based on sample entropy under RSS. ; and MSRSS designs. A Monte Carlo simulation study is carried out to compare the power of the proposed tests under several alternative distributions with the ordinary test based on SRS. The results report that the new entropy tests have higher power than the original one for nearly all sample sizes and under alternatives considered. Mahdizadeh, MahdiMon, 11 Nov 2013 14:25:40 GMThttp://ddd.uab.cat/record/113241Flexible geoadditive survival analysis of non-Hodgkin lymphoma in Peru
http://ddd.uab.cat/record/102780
Knowledge of prognostic factors is an important task for the clinical management of Non Hodgkin Lymphoma (NHL). In this work, we study the variables affecting survival of NHL in Peru by means of geoadditive Cox-type structured hazard regression models while accounting for potential spatial correlations in the survival times. We identified eight covariates with significant effect for overall survival. Some of them are widely known such as age, performance status, clinical stage and lactic dehydrogenase, but we also identified hemoglobin, leukocytes and lymphocytes as covariates with a significant effect on the overall survival of patients with NHL. Besides, the effect of continuous covariates is clearly nonlinear and hence impossible to detect with the classical Cox method. Although the spatial component does not show a significant effect, the results show a trend of low risk in certain areas. Flores, ClaudioWed, 09 Jan 2013 19:12:34 GMThttp://ddd.uab.cat/record/102780Statistical modelling and forecasting of outstanding liabilities in non-life insurance
http://ddd.uab.cat/record/102779
Non-life insurance companies need to build reserves to meet their claims liability cash flows. They often work with aggregated data. Recently it has been suggested that better statistical properties can be obtained when more aggregated data are available for statistical analysis than just the classical aggregated payments. When also the aggregated number of claims is available one can define a full statistical model of the nature of the number of claims, their delay until payment and the nature of these payments. In this paper we provide a new development in this direction by entering yet another set of aggregated data, namely the number of payments and when they occurred. A new element of our statistical analysis is that we are able to incorporate inflationary trends of payments in a direct and explicit way. Our new method is illustrated on a real life data set. Martínez-Miranda, María DoloresWed, 09 Jan 2013 18:59:24 GMThttp://ddd.uab.cat/record/102779Hurdle negative binomial regression model with right censored count data
http://ddd.uab.cat/record/102777
A Poisson model typically is assumed for count data. In many cases because of many zeros in the response variable, the mean is not equal to the variance value of the dependent variable. Therefore, the Poisson model is no longer suitable for this kind of data. Thus, we suggest using a hurdle negative binomial regression model to overcome the problem of overdispersion. Furthermore, the response variable in such cases is censored for some values. In this paper, a censored hurdle negative binomial regression model is introduced on count data with many zeros. The estimation of regression parameters using maximum likelihood is discussed and the goodness-of-fit for the regression model is examined. Ehsan Saffari, SeyedWed, 09 Jan 2013 18:48:03 GMThttp://ddd.uab.cat/record/102777The new class of Kummer beta generalized distributions
http://ddd.uab.cat/record/102776
Ng and Kotz (1995) introduced a distribution that provides greater flexibility to extremes. We define and study a new class of distributions called the Kummer beta generalized family to extend the normal, Weibull, gamma and Gumbel distributions, among several other well-known distributions. Some special models are discussed. The ordinary moments of any distribution in the new family can be expressed as linear functions of probability weighted moments of the baseline distribution. We examine the asymptotic distributions of the extreme values. We derive the density function of the order statistics, mean absolute deviations and entropies. We use maximum likelihood estimation to fit the distributions in the new class and illustrate its potentiality with an application to a real data set. Pescim, R. R.Wed, 09 Jan 2013 18:29:59 GMThttp://ddd.uab.cat/record/102776A note on the use of supply-use tables in impact analyses
http://ddd.uab.cat/record/102772
Little attention has so far been paid to the problems inherent in interpreting the meaning of results from standard impact analyses using symmetric input-output tables. Impacts as well as drivers of these impacts must be either of the product type or of the industry type. Interestingly, since supply-use tables distinguish products and industries, they can cope with product impacts driven by changes in industries, and vice versa. This paper contributes in two ways. Firstly, the demanddriven Leontief quantity model, both for industry-by-industry as well as for product-by-product tables, is formalised on the basis of supply-use tables, thus leading to impact multipliers, both for industries and products. Secondly, we demonstrate how the supply-use formulation can improve the incorporation of disparate satellite data into input-output models, by offering both industry and product representation. Supply-use blocks can accept any mix of industry and product satellite data, as long as these are not overlapping. Lenzen, ManfredWed, 09 Jan 2013 17:41:47 GMThttp://ddd.uab.cat/record/102772On developing ridge regression parameters : a graphical investigation
http://ddd.uab.cat/record/102770
In this paper we review some existing and propose some new estimators for estimating the ridge parameter. All in all 19 different estimators have been studied. The investigation has been carried out using Monte Carlo simulations. A large number of different models have been investigated where the variance of the random error, the number of variables included in the model, the correlations among the explanatory variables, the sample size and the unknown coefficient vector were varied. For each model we have performed 2000 replications and presented the results both in term of figures and tables. Based on the simulation study, we found that increasing the number of correlated variable, the variance of the random error and increasing the correlation between the independent variables have negative effect on the mean squared error. When the sample size increases the mean squared error decreases even when the correlation between the independent variables and the variance of the random error are large. In all situations, the proposed estimators have smaller mean squared error than the ordinary least squares and other existing estimators. Muniz, GiselaWed, 09 Jan 2013 17:37:29 GMThttp://ddd.uab.cat/record/102770Multiplicative noise for masking numerical microdata with constraints
http://ddd.uab.cat/record/97850
Before releasing databases which contain sensitive information about individuals, statistical agencies have to apply Statistical Disclosure Limitation (SDL) methods to such data. The goal of these methods is to minimize the risk of disclosure of the confidential information and at the same time provide legitimate data users with accurate information about the population of interest. SDL methods applicable to the microdata (i. e. collection of individual records) are often called masking methods. In this paper, several multiplicative noise masking schemes are presented. These schemes are designed to preserve positivity and inequality constraints in the data together with the vector of means and covariance matrix. Oganian, AnnaThu, 26 Jul 2012 13:39:04 GMThttp://ddd.uab.cat/record/97850