STATISTICS

Author: ANDERIZ LOPEZ MIGUEL.
Year: 2004.
University: NAVARRA.
Place of defense: UNIVERSIDAD PUBLICA DE NAVARRA.
Place of preparation: UNIVERSIDAD PUBLICA DE NAVARRA.

Summary: The work focuses exclusively on professionals in the health sciences, especially médicos.El author questions the basic assumptions as fact lque culture statistics of these professionals is no different from graduates in disciplines not mathematics or in those that are not study statistics as a subject essential. Such professionals need to acquire basic knowledge of statistics to improve their situation profesional.Ante the difficulty for them the purchase or memory of the foundation in math which is based statistics, prefiern learn to use statistical packages of computer, especially the SPSS. The working theory is developed on two supports: the author's experience of over 25 years teaching professionals statistical health sciences and a response form anonymously, which consists of two parts: an opinion survey and a questionnaire of statistical knowledge simple. It gets 56 responses that are dealt with by normal procedures, including factor analysis correlation. Are doctors participating in a level of statistical knowledge mid-level low, with committing the same mistakes that other professionals, and even university students and teaching media.La receipt of a statistical theoretical and practical course of five credits, taught for three weeks, in a very obvious improvement that level of knowledge. They are influential factors on the quality of responses to the questionnaire: date licenciatrura between 5 and 15 years ago, the number of hours pstgrado devoted to receive lessons on statistics, number of publications and the English proficiency level of translating with sultura scientific articles in professional journals. It also conducted several workshops in which proves that there is no significant difference between the statistical knowledge of the graduates and graduates in the various fields of health sciences: nursing, social graduates, and so on. It includes, as an annex to the work, the text of current statistics for health sciences to have made up reference.

Author: CASTRO LOPEZ CLAUDIO RAFAEL.
Year: 2004.
University: SALAMANCA.
Place of defense: FACULTAD DE MEDICINA.
Place of preparation: ESTADISTICA.

Summary: In the context of the family of segmentation methods AID (Automatic Interaction Detection), the best known of them is CHAID (CHI-square AID). CHAID uses a set of explanatory variables and a predictor variable, all categorical. The aim of the method is to segment the population or sample into groups of individuals as homogeneous as possible, for the variable response. In different stages of the method using the Chi-square test, for example, in phase grouping of categories of each predictor variable and the choice of the best predictor with respect to the dependent variable (first two phases of the algorithm). Employing the test Chi wouldnt be introduced limiting a procedure that is supposed asymmetric in its approach. That is, using the Chi-square test did not capture the asymmetric nature of a variable response and a predictor in the contingency table, it finds that the variable response can be ordinal scale, does not verify conditions colapsabilidad variables in the multidimensional table one, is not sensitive to detect the phenomenon known as Paradox Simpson, and carries a large number of hypothesis tests could present a risk of Type I error. Faced with this problem, the working thesis develops a matching algorithm to the phase of colapsamiento categories, using a model effects column, to capture the underlying order in the categories of the variable response. It deepens the colapsamiento variables using graphical models for contingency tables with multivariate models and response structures in blocks chained, as these models are behind a process of segmentation. It develops a method for segmentation based on obtaining a latent variable (not observable), which is obtained through a model of latent classes, from set of variables apparent response, this procedure allows us to consider issues in the multivariate variable response. The method developed finds perform Analysis No. Symmetrical the contingency table, which is formed between the predictor variable rate predictividad more with respect to the variable response. The segmentation is based on the estimated scores (coordinates) of the predictor variable categories, identifying in this way categories with strong and weak predictividad, same as giving rise to obtaining homogeneous segments over a variable latency response. It shows that the segments derived are highly predictive variables obvious why segmentation is common to a set of variables response.

Author: LOPEZ PINTADO SARA.
Year: 2004.
University: CARLOS III DE MADRID.
Place of defense: FACULTAD DE CIENCIAS SOCIALES Y JURIDICAS.
Place of preparation: UNIVERSIDAD CARLOS III DE MADRID.

Summary: The spread of new technologies and the growing complexity of the statistical analysis in various disciplines, such as economics, biology or medicine, often generate data in the form of duties. This thesis proposes a methodology for analyzing functional data based on the idea of depth. First, it introduces a number of definitions of the notion of depth for functional data are analyzed and their properties. The version finito-dimensional of these new concepts provides an alternative to all existing notions of depth that is computationally feasible in any dimension, and therefore suitable for any type of observations of great complexity. In addition, functional data to extend the ideas of cut and central regions are studied and their properties. It also presents strategies contrasting scenarios based on the new definitions to decide whether two sets of curves from the same population. Finally, are built sturdy supervised classification procedures and apply to data from microarrays.

Author: VEGAS AZCÁRATE SUSANA.
Year: 2005.
University: REY JUAN CARLOS.
Place of defense: ESCUELA SUPERIOR DE CIENCIAS EXPERIMENTALES Y TECNOLOGIA.
Place of preparation: ESCUELA SUPERIOR DE CIENCIAS EXPERIMENTALES Y TECNOLOGÍAS.

Summary: This thesis discusses a variety of topics related to the use of the structures of maps auto-organizativos for detecting multivariate fashions and the estimated densities. There are many ways to train these maps and guides little about what kind of map that kind of training regimen is most useful for each task. This thesis seeks to make up for this lack reviewing an important part of the literarura more relevant and to provide innovative solutions to some specific and known problems.

Author: HIERRO FRANCO MARIA.
Year: 2005.
University: CANTABRIA.
Place of defense: FACULTAD DE CIENCIAS ECONOMICAS Y EMPRESARIALES.
Place of preparation: FACULTAD DE CIENCIAS ECONOMICAS Y EMPRESARIALES.

Summary: This thesis proposes a general procedure for modeling dynamic internal migrations, which incorporates a number of significant developments with regard to the model of parent causative constant: the model of parent causative variables. The application of this model parent causative constant: the model of parent causative variables. The application of this model to the system of internal migration in Spain at the regional level for the period 1986-2003 can be drawn from a number of findings related to stability, convergence, degree of mobility between regions and identification of areas with greater power and withholding of deportation population. It was also confirmed, in accordance with the pattern of behavior followed by migration rates, the influence of cilo economic mobility between the degree and obtaining by the Spanish regions of gains or losses in the power of attraction or expulsion.

Author: MARTINEZ MARTINEZ JOSE MIGUEL.
Year: 2005.
University: POLITÉCNICA DE CATALUÑA.
Place of defense: Universidad Politécnica de Catalunya.
Place of preparation: OMEGA DESPATX 407, PLANTA 4º NORD.

Author: GOICOA MANGADO TOMÁS.
Year: 2005.
University: PÚBLICA DE NAVARRA.
Place of defense: ESCUELA TÉCNICA SUPERIOR DE INGENIEROS INDUSTRIALES Y DE TELECOMUNICACIÓN.
Place of preparation: ESCUELA TÉCNICA SUPERIOR DE INGENIEROS INDUSTRIALES Y DE TELECOMUNICACIÓN.

Author: SERRA CUÑAT JUAN FRANCISCO.
Year: 2005.
University: COMPLUTENSE DE MADRID.
Place of defense: FACULTAD DE CIENCIAS MATEMÁTICAS.
Place of preparation: FACULTAD DE CIENCIS MATEMÁTICAS (UCM).

Summary: This paper deals with the study of entropy in an experiment, which shows a discrete random variable whose probability distribution is a linear transformation of a space initial probability finite 0. It proposes a general method for obtaining the distribution 0 leading to experiment maximum entropy, or one close to maximum entropy, in the transformed space. This distribution depends on the extent of entropy chosen. We present an application to the analysis of survival when considering an experiment randomly censored by the right, with data grouped at intervals, such as that resulting from a transformation of the corresponding linear experiment is not censored. The results are illustrated with several examples on different distributions of censorship and different measure of entropy.

Author: CORREA ESPINAL ALEXANDER ALBERTO.
Year: 2006.
University: POLITÉCNICA DE CATALUÑA.
Place of defense: Facultat de Matemàtiques i Estadística.
Place of preparation: ETSEIB, Edifici H PLANTA 6, DESPATX: 6.63 SD.

Summary: A common recommendation when thinking in a factorial design is randomizing the run order. The purpose of this randomization is to protect the response from the possible influence of unknown factors. This influence is expected to be blurred among all the effects, thus none of them is specially affected and no mistakes are made when estimating its statistical significance. But this praxis has two essential problems: 1. The number of factorâs level changes due to randomization might be large (much larger than in other sequences). It can be also difficult to conduct, making the experimentation complicated and more expensive. 2. Making some reasonable hypothesis regarding the influence of the unknown factors, there are some sequences clearly better than others for minimizing the influence of this undesirable factors. Many authors have worked on this topic, and some matters have already been solved. For instance, the experimentation sequence that better neutralises the influence of unknown factors is already determined, but without taking into consideration the number of level changes that this sequence implies. It has also been solved the problem of finding sequences that have the minimum number of level changes, but without considering simultaneously the potential influence of unknown factors. When both the influence of unknown factors and the number of level changes is considered, the problem has been solved up to designs with 16 runs. But not further as the searching procedures used are nonviable when the number of posible sequences becomes so huge(with 32 runs the number of different sequences is 32!=2,6·1035) The aim of this thesis is finding a procedure that makes it possible to obtain run sequences with the minimum number of level changes, and that besides minimize the influence of unknown factors in the effect estimation, for any 2 level factorial design. Moreover, the desired run sequence should be obtained easily by the experimenter when using the proposed procedure. The content is structured in 7 chapters and 8 appendixes. Chapter 1 shows the motivation that lead to chose this research topic. It also defines the basic elements of this work (complete and fractional 2 level factorial designs, problems that appear when randomizing this designs, and how to quantify the influence of unknown and undesired factors in the effect estimation). In addition, the hypothesis and context in which the search for run orders with the desired properties will take place are presented. Chapter 2 gives an exhaustive bibliographic review of the current solutions related with run orders in these designs robust to the influence of factors alien to the experimentation and/or with minimum number of level changes. The end of the chapter lists weaknesses of the current state of the art and advances the expected contributions of this thesis. Chapter 3 presents an original procedure for finding run orders for 2 level factorial designs with the minimum number of changes in the level factors and a known bias. We called this procedure duplication method, as duplicating the rows of a 2k design and adding a factor with a specific sign sequence, a 2k+1 design with the same properties as the first design is achieved. An important property of this method is that it can be applied to any number of factors. This procedure guarantees the minimum number of level changes, but not always guaranties the minimum bias (measure of the influence that unknown factors have in the effect estimation). Chapter 4 shows different methods for finding run orders with less bias than the one produced by the duplication method. These methods are: - Random search with restrictions: The procedure randomly generates the run order, but in a way that a run is followed by another one that has only one change in the factor levels (the minimum number of changes is then guaranteed). Once the sequence is completed its bias is calculated.

Author: CINTAS DEL RÍO ROSARIO.
Year: 2006.
University: COMPLUTENSE DE MADRID.
Place of defense: FACULTAD DE CIENCIAS MATEMÁTICAS.
Place of preparation: FACULTAD DE CIENCIAS MATEMÁTICAS - UCM.

Summary: The thesis presents and solves the problem of adapting and create models based on functions mating, both static and dynamic, able to capture relevant traits series financiares bivariate, in order to be useful in monitoring and assessing potential risks of financial markets . They apply these models to analyze the behavior of losses joint financial indices Dow Jones and Ibex35, as reflected in the lower tail of the bivariate distribution. The results are significant both in terms of theoretical and practical.

Author: VÉLEZ SERRANO DANIEL.
Year: 2006.
University: COMPLUTENSE DE MADRID.
Place of defense: FACULTAD DE CIENCIAS MATEMÁTICAS.
Place of preparation: FACULTAD DE CIENCIAS MATEMÁTICAS - UCM.

Summary: It is proposed to create mating model for the application to the predictions and considered measures of dependence between two variables and establishing mating reflecting such dependence. These are used mating to a specific problem of predicting peak daily energy demand (gas or electricity in the short and medium term. The thesis is divided into seven chapters and an appendix. Also involves an innovation in the use of mating in the prediction .