Essays on mixed-frequency data : forecasting and unit root testing

Doutoramento em Economia

Um novo horizonte para a cooperação bilateral entre Espanha e Portugal em matéria de defesa

Os Estados Membros da UE devem assumir a responsabilidade pela sua defesa e segurança, num novo quadro com menos presença dos EUA e novas ameaças. A perda de capacidades militares desde o início da crise é evidente, tornando-se necessária uma partilha que permita manter a eficiência e a economia de meios. Nesta conjuntura, Portugal e Espanha têm definido no Tratado de Baiona (2015) o mais ambicioso quadro legal da sua história para uma necessária cooperação bilateral em defesa. Uma iniciativa que pode e deve servir de primeiro passo na integração da defesa comum europeia. O presente trabalho desenha uma estratégia de cooperação bilateral, materializada num modelo de cooperação construído sobre o Tratado de Baiona, que resulta de analisar os fins, os meios e os modos de cooperação em europa nos níveis regional (OTAN e UE) e sub-regional (bilaterais e minilaterais). Abstract: The Member States of the UE need to assume the responsibility of their defense and security, in the framework of a reduced presence of the USA and increased threats. The loss of military capabilities from the beginning of the crises appears to be evident, making necessary the establishment of a sharing procedure that ensures efficiency and economy of means. Within this situation, Portugal and Spain have defined in the Treaty of Baiona (2015) the most ambitious legal framework ever in their common history, for an enhanced bilateral defense cooperation. This initiative may and must serve as first step in the integration of an European common defense. The present work designs a strategy for bilateral cooperation, materialized in a cooperation model build upon the Treaty of Baiona, as a result of the analysis of the ends, ways and means of the European cooperation at both regional (NATO & EU) and sub-regional (bilateral & minilateral) levels.

Relevance or irrelevance of capital structure?

In this paper we examine the effects of asymmetric information on the nature of financial equilibrium and on the capital structure of firms. In the first model presented, the financial contracts on offer involve pooling equilibrium with no adverse selection. However, in the special case analyzed, where contracts are of mixed form, there may be a separating equilibrium and also equilibrium may not exist. Interesting result is that the separating equilibrium found is not economically efficient since aggregate investments falls short of first-best level. More importantly, capital structure does matter. The relative magnitude of outside equity makes a real difference to the quantity of aggregate investment in equilibrium.

Modelo de gestión para la atención integral de pacientes con enfermedades huérfanas o raras en Colombia

Las enfermedades raras o huérfanas corresponden a aquellas con baja prevalencia en la población, y en varios países tienen una definición distinta de acuerdo con el número de pacientes que afectan en la población. La Organización Mundial de la Salud (OMS), las define como un trastorno que afecta de 650 a 1.000 personas por millón de habitantes, de las que se han identificado alrededor de 7.000. En Colombia su prevalencia es menor de 1 por cada 5.000 personas y comprenden: las enfermedades raras, las ultra-huérfanas y las olvidadas. Los pacientes con este tipo de enfermedades imponen retos a los sistemas sanitarios, pues si bien afectan a un bajo porcentaje de la población, su atención implica una alta carga económica por los costos que involucra su atención, la complejidad en su diagnóstico, tratamiento, seguimiento y rehabilitación. El abordaje de las enfermedades raras requiere un manejo interdisciplinar e intersectorial, lo que implica la organización de cada actor del sistema sanitario para su manejo a través de un modelo que abraque las dinámicas posibles entre ellos y las competencias de cada uno. Por lo anterior, y teniendo en cuenta la necesidad de formular políticas sanitarias específicas para la gestión de estas enfermedades, el presente trabajo presenta una aproximación a la formulación de un modelo de gestión para la atención integral de pacientes con enfermedades raras en Colombia. Esta investigación describe los distintos elementos y características de los modelos de gestión clínica y de las enfermedades raras a través de una revisión de literatura, en la que se incluye la descripción de los distintos actores del Sistema de Salud Colombiano, relacionados con la atención integral de estos pacientes para la documentación de un modelo de gestión integral.

Frequency domain analysis of stationary time series

This thesis provides a necessary and sufficient condition for asymptotic efficiency of a nonparametric estimator of the generalised autocovariance function of a Gaussian stationary random process. The generalised autocovariance function is the inverse Fourier transform of a power transformation of the spectral density, and encompasses the traditional and inverse autocovariance functions. Its nonparametric estimator is based on the inverse discrete Fourier transform of the same power transformation of the pooled periodogram. The general result is then applied to the class of Gaussian stationary ARMA processes and its implications are discussed. We illustrate that for a class of contrast functionals and spectral densities, the minimum contrast estimator of the spectral density satisfies a Yule-Walker system of equations in the generalised autocovariance estimator. Selection of the pooling parameter, which characterizes the nonparametric estimator of the generalised autocovariance, controlling its resolution, is addressed by using a multiplicative periodogram bootstrap to estimate the finite-sample distribution of the estimator. A multivariate extension of recently introduced spectral models for univariate time series is considered, and an algorithm for the coefficients of a power transformation of matrix polynomials is derived, which allows to obtain the Wold coefficients from the matrix coefficients characterizing the generalised matrix cepstral models. This algorithm also allows the definition of the matrix variance profile, providing important quantities for vector time series analysis. A nonparametric estimator based on a transformation of the smoothed periodogram is proposed for estimation of the matrix variance profile.

Tecniche di group testing per il rilevamento di SARS-CoV-2

Oggigiorno l'individuazione diagnostica precoce della SARS-CoV-2 attraverso tamponi molecolari è fondamentale per interrompere la trasmissione del virus. Tuttavia, il monitoraggio della diffusione virale attraverso il test standard di RT-qPCR individuale comporta un elevato costo per ciascun tampone nasofaringeo analizzato e i reagenti chimici per l’estrazione dell’RNA virale sono sempre meno disponibili. Per ovviare a tali ostacoli, è stata ripresa la tecnica di group testing, sviluppata per la prima volta da Dorfman nel 1943 per individuare i soggetti affetti da sifilide prima del loro arruolamento. Questa strategia minimizza il numero di test condotti su un insieme di campioni: se un gruppo di n campioni risulta negativo, allora la condizione di ciascuno di essi è stata determinata mediante un solo test invece che con n test individuali. Negli ultimi due anni sono state sviluppate strategie in grado di migliorare le prestazioni del test di gruppo: per scenari a bassa prevalenza l’algoritmo dell’ipercubo rileva un singolo campione positivo in pool con dimensioni fino a 100 campioni attraverso due o più turni di test; invece, il P-BEST utilizza un solo turno di analisi, ma le dimensioni massime dei pool sono più ridotte. Per scenari ad alta prevalenza (10%) il team italiano dell’Università di Bologna ha progettato un metodo che identifica e rileva i membri infetti con un solo turno di test. Tuttavia, affinché il group testing sia efficace come l’analisi individuale dei tamponi molecolari, è necessario minimizzare l’effetto di diluizione, correlato alla dimensione del pool e causa di insorgenza di falsi negativi, nonché di un calo nella sensibilità nei test. I dati ottenuti da questi studi hanno dimostrato che questa strategia offre grandi potenzialità. Essa è essenziale per le indagini di routine della popolazione e concede vantaggi amplificati soprattutto se vengono testati soggetti quotidianamente in contatto tra loro, come famiglie o colleghi di lavoro.

Geometric deep learning for polygonal mesh denoising

Il seguente lavoro si propone come analisi degli operatori convoluzionali che caratterizzano le graph neural networks. ln particolare, la trattazione si divide in due parti, una teorica e una sperimentale. Nella parte teorica vengono innanzitutto introdotte le nozioni preliminari di mesh e convoluzione su mesh. In seguito vengono riportati i concetti base del geometric deep learning, quali le definizioni degli operatori convoluzionali e di pooling e unpooling. Un'attenzione particolare è stata data all'architettura Graph U-Net. La parte sperimentare riguarda l'applicazione delle reti neurali e l'analisi degli operatori convoluzionali applicati al denoising di superfici perturbate a causa di misurazioni imperfette effettuate da scanner 3D.

Design and management of a supply chain: a joint location-inventory problem under uncertainty (Part B)

In recent years, global supply chains have increasingly suffered from reliability issues due to various external and difficult to-manage events. The following paper aims to build an integrated approach for the design of a Supply Chain under the risk of disruption and demand fluctuation. The study is divided in two parts: a mathematical optimization model, to identify the optimal design and assignments customer-facility, and a discrete-events simulation of the resulting network. The first one describes a model in which plant location decisions are influenced by variables such as distance to customers, investments needed to open plants and centralization phenomena that help contain the risk of demand variability (Risk Pooling). The entire model has been built with a proactive approach to manage the risk of disruptions assigning to each customer two types of open facilities: one that will serve it under normal conditions and a back-up facility, which comes into operation when the main facility has failed. The study is conducted on a relatively small number of instances due to the computational complexity, a matheuristic approach can be found in part A of the paper to evaluate the problem with a larger set of players. Once the network is built, a discrete events Supply Chain simulation (SCS) has been implemented to analyze the stock flow within the facilities warehouses, the actual impact of disruptions and the role of the back-up facilities which suffer a great stress on their inventory due to a large increase in demand caused by the disruptions. Therefore, simulation follows a reactive approach, in which customers are redistributed among facilities according to the interruptions that may occur in the system and to the assignments deriving from the design model. Lastly, the most important results of the study will be reported, analyzing the role of lead time in a reactive approach for the occurrence of disruptions and comparing the two models in terms of costs.