Nel corso degli ultimi venti anni si è assistito ad un incremento generalizzato della disoccupazione in Europa in particolare di quella giovanile: nel 1997 in molti paesi europei il tasso di disoccupazione della classe 15-24 anni è doppio rispetto a quello degli adulti. Questo lavoro si propone di dare una descrizione realistica del fenomeno della disoccupazione giovanile in Italia come risultante dalle prime 4 waves dell’indagine European Household Community Panel indagando la probabilità di transizione dallo stato di disoccupazione a quello di occupazione. Nell’ambito di un’impostazione legata alla teoria dei processi stocastici e dei dati di durata si indagheranno gli effetti che episodi di disoccupazione precedenti possono avere sulla probabilità di trovare un lavoro, in particolare, nell’ambito di processi stocastici più generali si rilascerà l’ipotesi di semi-markovianità del processo per considerare l’effetto di una funzione della storia passata del processo sulla transizione attuale al lavoro. La stima della funzione di rischio a vari intervalli di durata si dimostra più appropriata quando si tiene conto della storia passata del processo, in particolare, si verifica l’ipotesi che la possibilità di avere successo nella ricerca di un lavoro è negativamente influenzata dall’aver avuto in passato molte transizioni disoccupazione-occupazione-disoccupazione.

The aim of this work is to carry out an applicative, comparative and exhaustive study between several entropy based indicators of independence and correlation. We considered some indicators characterized by a wide and consolidate literature, like mutual information, joint entropy, relative entropy or Kullback Leibler distance, and others, more recently introduced, like Granger, Maasoumi and racine entropy, also called Sρ, or utilized in more restricted domains, like Pincus approximate entropy or ApEn. We studied the behaviour of such indicators applying them to binary series. The series was designed to simulate a wide range of situations in order to characterize indicators limit and capability and to identify, case by case, the more useful and trustworthy ones. Our target was not only to study if such indicators were able to discriminate between dependence and independence because, especially for mutual information and Granger, Maasoumi and Racine, that was already demonstrated and reported in literature, but also to verify if and how they were able to provide information about structure, complexity and disorder of the series they were applied to. Special attention was paid on Pincus approximate entropy, that is said by the author to be able to provide information regarding the level of randomness, regularity and complexity of a series. By means of a focused and extensive research, we furthermore tried to clear the meaning of ApEn applied to a couple of different series. In such situation the indicator is named in literature as cross-ApEn. The cross-ApEn meaning and the interpretation of its results is often not simple nor univocal and the matter is scarcely delved into by literature, thereby users can easily leaded up to a misleading conclusion, especially if the indicator is employed, as often unfortunately it happens, in uncritical manner. In order to plug some cross-ApEn gaps and limits clearly brought out during the experimentation, we developed and applied to the already considered cases a further indicator we called “correspondence index”. The correspondence index is perfectly integrated into the cross-ApEn computational algorithm and it is able to provide, at least for binary data, accurate information about the intensity and the direction of an eventual correlation, even not linear, existing between two different series allowing, in the meanwhile, to detect an eventual condition of independence between the series themselves.

We propose an extension of the approach provided by Kluppelberg and Kuhn (2009) for inference on second-order structure moments. As in Kluppelberg and Kuhn (2009) we adopt a copula-based approach instead of assuming normal distribution for the variables, thus relaxing the equality in distribution assumption. A new copula-based estimator for structure moments is investigated. The methodology provided by Kluppelberg and Kuhn (2009) is also extended considering the copulas associated with the family of Eyraud-Farlie-Gumbel-Morgenstern distribution functions (Kotz, Balakrishnan, and Johnson, 2000, Equation 44.73). Finally, a comprehensive simulation study and an application to real financial data are performed in order to compare the different approaches.

In a large number of problems the high dimensionality of the search space, the vast number of variables and the economical constrains limit the ability of classical techniques to reach the optimum of a function, known or unknown. In this thesis we investigate the possibility to combine approaches from advanced statistics and optimization algorithms in such a way to better explore the combinatorial search space and to increase the performance of the approaches. To this purpose we propose two methods: (i) Model Based Ant Colony Design and (ii) Naïve Bayes Ant Colony Optimization. We test the performance of the two proposed solutions on a simulation study and we apply the novel techniques on an appplication in the field of Enzyme Engineering and Design.

The primary aim of this dissertation to identify subgroups of patients with chronic kidney disease (CKD) who have a differential risk of progression of illness and the secondary aim is compare 2 equations to estimate the glomerular filtration rate (GFR). To this purpose, the PIRP (Prevention of Progressive Kidney Disease) registry was linked with the dialysis and mortality registries. The outcome of interest is the mean annual variation of GFR, estimated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation. A decision tree model was used to subtype CKD patients, based on the non-parametric procedure CHAID (Chi-squared Automatic Interaction Detector). The independent variables of the model include gender, age, diabetes, hypertension, cardiac diseases, body mass index, baseline serum creatinine, haemoglobin, proteinuria, LDL cholesterol, tryglycerides, serum phoshates, glycemia, parathyroid hormone and uricemia. The decision tree model classified patients into 10 terminal nodes using 6 variables (gender, age, proteinuria, diabetes, serum phosphates and ischemic cardiac disease) that predict a differential progression of kidney disease. Specifically, age <=53 year, male gender, proteinuria, diabetes and serum phosphates >3.70 mg/dl predict a faster decrease of GFR, while ischemic cardiac disease predicts a slower decrease. The comparison between GFR estimates obtained using MDRD4 and CKD-EPI equations shows a high percentage agreement (>90%), with modest discrepancies for high and low age and serum creatinine levels. The study results underscore the need for a tight follow-up schedule in patients with age <53, and of patients aged 54 to 67 with diabetes, to try to slow down the progression of the disease. The result also emphasize the effective management of patients aged>67, in whom the estimated decrease in glomerular filtration rate corresponds with the physiological decrease observed in the absence of kidney disease, except for the subgroup of patients with proteinuria, in whom the GFR decline is more pronounced.

In the present work we perform an econometric analysis of the Tribal art market. To this aim, we use a unique and original database that includes information on Tribal art market auctions worldwide from 1998 to 2011. In Literature, art prices are modelled through the hedonic regression model, a classic fixed-effect model. The main drawback of the hedonic approach is the large number of parameters, since, in general, art data include many categorical variables. In this work, we propose a multilevel model for the analysis of Tribal art prices that takes into account the influence of time on artwork prices. In fact, it is natural to assume that time exerts an influence over the price dynamics in various ways. Nevertheless, since the set of objects change at every auction date, we do not have repeated measurements of the same items over time. Hence, the dataset does not constitute a proper panel; rather, it has a two-level structure in that items, level-1 units, are grouped in time points, level-2 units. The main theoretical contribution is the extension of classical multilevel models to cope with the case described above. In particular, we introduce a model with time dependent random effects at the second level. We propose a novel specification of the model, derive the maximum likelihood estimators and implement them through the E-M algorithm. We test the finite sample properties of the estimators and the validity of the own-written R-code by means of a simulation study. Finally, we show that the new model improves considerably the fit of the Tribal art data with respect to both the hedonic regression model and the classic multilevel model.

Obiettivo: studio longitudinale di una coorte di gravidanze gemellari e valutazione di eventuali correlazioni tra specifici parametri cardiaci materni ed esito sfavorevole della gravidanza. Metodi: Sono state arruolate in modo prospettico donne con gravidanze gemellari, sottoposte a controlli seriati della funzione cardiaca materna ( 20-23 settimane; 26-29 settimane e 30-33 settimane). Le pazienti in cui il parto si è verificato prima della 34°settimana compiuta di gestazione sono state escluse a priori dallo studio. Specifici parametri cardiaci campionati sono stati confrontati tra il gruppo di gravidanze con esito positivo ed il gruppo di pazienti che hanno sviluppato complicanze nel corso della gravidanza quali: preeclampsia o ipertensione gestazionale, basso peso alla nascita ( SGA: peso neonatale di uno o entrambi i feti < 5° centile per l’epoca di gestazione). Risultati: sono state incluse nello studio 28 gravidanze gemellari,di cui 8 complicate. Durante ogni visita abbiamo osservato un aumento significativo della gittata cardiaca ( CO ) e sistolica ( SV ) ed una caduta delle resistenze vascolari totali ( TVR ) nelle gravidanze gemellari normali vs complicate. Inoltre, CO, pressione arteriosa (BP), frequenza cardiaca ( HR ) e TVR son rimaste invariate nel corso dei controlli ambulatoriali per le gravidanze con esiti sfavorevole, mentre mostravano modificazioni significative ( CO, HR e BDP in aumento; TVR in diminuzione ) nel gruppo di pazienti con esito favorevole della gravidanza. Conclusioni: La valutazione longitudinale della funzione cardiaca materna sembra dimostrare una significativa differenza tra i valori dei parametri cardiaci materni delle gravidanze multiple non complicate, rispetto a quelle complicate da preeclampsia o neonati SGA. / Objective: to assess if the occurrence of adverse pregnancy outcome may be related to specific maternal cardiac parameters. Study design: women with twin pregnancies were prospectively submitted to serial maternal echocardiography (20-23 weeks; 26-29 weeks and 30-33 weeks). Cardiac findings were compared between patients with uneventful pregnancy outcome and those who developed one of the following complications: preeclampsia or gestational hypertension, SGA neonates or IUGR. Results: A group of 28 twin gestations was obtained for analysis, including 8 complicated cases. At each visit cardiac output (CO) and stroke volume (SV) were significantly higher whereas total vascular resistance (TVR) was lower among uncomplicated vs complicated twin pregnancies. Conclusions: At longitudinal evaluation, maternal cardiac findings seem to be significantly different between uneventful twin pregnancies and those complicated.

In the last couple of decades we assisted to a reappraisal of spatial design-based techniques. Usually the spatial information regarding the spatial location of the individuals of a population has been used to develop efficient sampling designs. This thesis aims at offering a new technique for both inference on individual values and global population values able to employ the spatial information available before sampling at estimation level by rewriting a deterministic interpolator under a design-based framework. The achieved point estimator of the individual values is treated both in the case of finite spatial populations and continuous spatial domains, while the theory on the estimator of the population global value covers the finite population case only. A fairly broad simulation study compares the results of the point estimator with the simple random sampling without replacement estimator in predictive form and the kriging, which is the benchmark technique for inference on spatial data. The Monte Carlo experiment is carried out on populations generated according to different superpopulation methods in order to manage different aspects of the spatial structure. The simulation outcomes point out that the proposed point estimator has almost the same behaviour as the kriging predictor regardless of the parameters adopted for generating the populations, especially for low sampling fractions. Moreover, the use of the spatial information improves substantially design-based spatial inference on individual values.

The thesis is concerned with local trigonometric regression methods. The aim was to develop a method for extraction of cyclical components in time series. The main results of the thesis are the following. First, a generalization of the filter proposed by Christiano and Fitzgerald is furnished for the smoothing of ARIMA(p,d,q) process. Second, a local trigonometric filter is built, with its statistical properties. Third, they are discussed the convergence properties of trigonometric estimators, and the problem of choosing the order of the model. A large scale simulation experiment has been designed in order to assess the performance of the proposed models and methods. The results show that local trigonometric regression may be a useful tool for periodic time series analysis.

The aim of the thesis is to propose a Bayesian estimation through Markov chain Monte Carlo of multidimensional item response theory models for graded responses with complex structures and correlated traits. In particular, this work focuses on the multiunidimensional and the additive underlying latent structures, considering that the first one is widely used and represents a classical approach in multidimensional item response analysis, while the second one is able to reflect the complexity of real interactions between items and respondents. A simulation study is conducted to evaluate the parameter recovery for the proposed models under different conditions (sample size, test and subtest length, number of response categories, and correlation structure). The results show that the parameter recovery is particularly sensitive to the sample size, due to the model complexity and the high number of parameters to be estimated. For a sufficiently large sample size the parameters of the multiunidimensional and additive graded response models are well reproduced. The results are also affected by the trade-off between the number of items constituting the test and the number of item categories. An application of the proposed models on response data collected to investigate Romagna and San Marino residents' perceptions and attitudes towards the tourism industry is also presented.