1359 research outputs found
Sort by
Objective bayesian variable selection for censored data
Get PDFIn this thesis we study the problem of selecting a set of regressors when the response variable follows a parametric model (such as Weibull or lognormal) and observations are right censored. Under a Bayesian approach, the most widely used tools are the Bayes Factors (BFs) which are, however,
undefined when using improper priors. Some commonly used tools in literature, which solve the problem of indeterminacy in model selection, are the Intrinsic Bayes factor (IBF) and the Fractional Bayes factor (FBF). The two
proposals are not actual Bayes factors but it can be shown that they asymptotically tend to actual BFs calculated over particular priors called intrinsic and fractional priors, respectively. Each of them depends on the size of a
minimal training sample (MTS) and, in particular, the IBF also depends on the MTSs used. When working with censored data, it is not immediate to define a suitable MTS because the sample space of response variables must be fully explored when drawing MTSs, but only uncensored data are
actually relevant to train the improper prior into a proper posterior. In fact, an unweighted MTS consisting only of uncensored data may produce a serious bias in model selection. In order to overcome this problem, a sequential
MTS (SMTS) is used, leading to an increase in the number of possible MTSs as each one has random size. This prevents the use of the IBF for exploring large model spaces. In order to decrease the computational cost, while maintaining a behavior comparable to that of the IBF, we provide a
suitable definition of the FBF that gives results similar to the ones of the IBF calculated over the SMTSs. We first define the conditional FBF on a fraction proportional to the MTS size and, then, we show that the marginal
FBF (mFBF), obtained by averaging the conditional FBFs with respect to the probability distribution of the fraction, is consistent and provides also good results. Next, we recall the definition of intrinsic prior for the case
of the IBF and the definition of the fractional prior for the FBF and we calculate them in the case of the exponential model for right censored data. In general, when the censoring mechanism is unknown, it is not possible to
obtain these priors. Also another approach to the choice of the MTS, which consists in weighting the MTS by a suitable set of weights, is presented. In fact, we define the
Kaplan-Meier minimal training sample (KMMTS) which depends on the Kaplan-Meier estimator of the survival function and which contains only suitable weighted uncensored observations. This new proposal could be useful when the censoring percentage is not very high, and it allows faster
computations when the predictive distributions, calculated only over uncensored observations, can be obtained in closed-form. The new methodologies are validated by means of simulation studies and applications to real data
Studio delle problematiche delle infiltrazioni e perdite della diga di Medau Zirimilis (Sardegna SW)
Get PDFThe costruction earthfill dam Medau Zirimilis, in the territory of Siliqua (SW Sardinia), began in 1989 and ended in 1991. Over the years, a time series of experimental basins, showed several problems of leakage and filtration in dam body. In the present work, through the application of classical methods of hydrogeology and applied, we have studied the problems of infiltration and loss of earth dam of Medau Zirimilis. The research has tried to examine the causes of the existing problems and identify possible solutions. The work of investigation has been developed from the acquisition of a basic knowledge of the geological characteristics of the tax of the dam and its surroundings. The understanding of its technical and constructional features and behavior had the objective to increase the level of reservoir . For this purpose, during the research have been acquired experimental data on levels in various types of water wells and drainage of the dam. Through the elaboration and interpretation with GIS methodologies and software, the attempt of this research was to reconstruct the model of the aquifer and to identify possible preferential of flow paths.
Always with the purpose of identifying pathways of water, we have been carried out a study on the mantle of coating internal hangings of the bank barrage made in bituminous conglomerate, applying innovative methodologies, techniques by such as non-destructive surveys to infrared thermography
Productivity, markups, and trade liberalization: evidence from Mexican manufacturing industries
Get PDFIn order to evaluate the eects of any policy or answer economic relevant questions it is of primary importance to accurately quantify the variables and the parameters
that may be involved with the policy or the questions. Since production functions are a fundamental components of all economics, in most of the cases answering those
questions requires to estimate production function parameters. Nonetheless, production function estimation is challenging since optimal input choices are correlated with
rm-specic unobserved productivity shocks. In this contribution I apply a structural framework for estimating production function coecients explicitly controlling for productivity through an observable proxy (investment or intermediate inputs) using a rich panel dataset of Mexican manufacturing rms between 1984 and 1990. I further derive
rm-level markup estimates and use these estimates to evaluate the impact of the dramatic trade liberalization that took place in Mexico during the sample period on the
protability of domestic rms and exporters. My ndings emphasize the importance of obtaining consistent estimates in order to correctly assess dierences in technologies,
productivity, and market power among the Mexican rms
Molecular and genetic characterization, clinical evaluation and pilot study to assess the feasibility of a carrier screening for Crisponi syndrome in Sardinia
Get PDFCrisponi syndrome (CS) and cold-induced sweating syndrome type 1 (CISS1) share clinical characteristics, such as dysmorphic features, muscle contractions, scoliosis and cold-induced sweating, with CS patients showing a severe clinical course in infancy involving hyperthermia,
associated with death in most cases in the first years of life. Functional and clinical studies supported the fact that they represent manifestations of the same autosomal recessive disorder caused by mutations in the CRLF1 gene with different degrees of severity. Anyway the debate
about what term is more appropriate to indicate the syndrome is still open. During my project we expanded the mutational spectrum of CRLF1 in the syndrome and carried out
a meta-analysis of the literature for all the mutations described so far. To date, overall 42 distinct CRLF1 mutations found either in homozygosity or in compound heterozygosity have been found in CS/CISS1 patients. In particular we are describing 11 new mutations in addition to the 31 already reported in literature along with a detailed clinical phenotype and genotype/phenotype correlation. The highest prevalence is registered in Sardinia, Turkey and Spain. In Sardinia the evaluation of the allele frequency for two out of the three founder mutations estimated a carrier frequency of 1,4%
with an incidence of about 1 case per 20,700 newborns
Stochastic modeling of nonisothermal antisolvent crystallization processes
Get PDFIn this Thesis a stochastic approach to model antisolvent
crystallization processes is addressed. The motivations to choice a stochastic approach instead of a population balance modeling has been developed to find a simple and an alternative way to describe the evolution of the Crystal Size Distribution (CSD), without consider complex thermodynamic and kinetic aspects of the process.
An important parameter to consider in crystallization process is the shape of the CSD (in terms of variance) and the mean size of crystals in order to optimize the filtering of the final product and then increase the production. The crystallization processes considered in this Thesis are the antisolvent crystallization processes
used in particular when the solute is weakly temperature-sensitive and then a second solvent, properly called antisolvent, is added in the solution favoring the crystallization of the solute. In antisolvent
crystallization processes it is important the consumption of the second solvent added, in particular, optimizing the feed-rate and coupling the process in synergy with cooling crystallization in order to improve the production and the quality of the desired product. The stochastic approach used in this Thesis is based on the Fokker Plank Equation (FPE), which has allowed finding an analytical
solution of the model, with some assumptions, and obtaining an analytical model able to describe the evolution of the mean size of crystals and the variance of the CSD. This analytical solution has leaded to develop an analytical relationship between the evolution in time of the first two stochastic moments of the FPE, such as mean and variance, and the manipulated variables, such as antisolvent
feed-rate and temperature, obtaining as a result a map showing the asymptotic moments obtainable within a certain range of operating conditions. This Thesis also analyzes the physical-chemical aspects of the antisolvent crystallization processes, including the temperature effects, finding a strong influence onto the nucleation and growth rate of crystals due by the hydrogen bond strength between solventantisolvent molecules despite of the molecular interaction in a solvated system.The physical-chemical consideration concerning the antisolvent
crystallization processes allowed to better understand the influence of the second solvent added, consequently optimizing the choice for the proper antisolvent to use with a proper feed-rate and temperature profile, minimizing the energy consumptions, in order to obtain the desired product. The stochastic model and the physical-chemical considerations have been validated with experimental data performed in a laboratory scale crystallizer. The experimental samples have been analyzed using an optical microscope and then the images taken have been manually processed in order to obtain the experimental CSDs
Design of functional colloidal magnetic nanoparticles for biomedical applications
Get PDFNanostructured materials have been the focus of scientific interest in recent years, because of the peculiar properties displayed by matter in nano-size form. The interest has been growing steeply since possible applications in biomedical fields have shown to
be realistic. The interaction of magnetic nanoparticles with applied magnetic field gradients makes these particles attractive for their potential applications in biomedical
imaging, diagnostic, and therapy. In this context, a great advantage are the superparamagnetic properties that can be exhibited by the nanoparticles. They behave as paramagnetic centres with high magnetic moment at room temperature, as it arises from the coupling of many atomic spins. After eliminating the external magnetic field, the particles no longer show magnetic interaction, thus reducing the possibility of particle aggregation; this feature is especially important for their applications. Biomedical use of magnetic nanoparticles imposes their uniform dispersion and stability in the biological fluids; moreover their size range should easily permit cell internalization through pinocytosis or endocytosis. Therefore, surface modification, via coating or encapsulation, is widely employed to improve nanoparticles properties and
for immobilization of functional molecules. Magnetic nanoparticles have garnered widespread attention in recent years to develop and understand synthetic means to control their size, magnetic behaviour, and chemical reactivity. Simultaneously tuning surface chemistry and physical properties enable preparation of functional magnetic
nanoparticles. Recent advances in synthesis have allowed to easily prepare a wide range of magnetic nanoparticles though aqueous or non aqueous approaches. It has been
widely shown that the non-aqueous routes are more efficient in producing stable colloidal nanoparticles with narrow size distribution, high crystallinity, tunable size and
shape. However, this approach typically produces hydrophobic nanoparticles limiting their applications in biological and medical fields. Thus, the transformation of these hydrophobic nanoparticles into hydrophilic is a crucial step toward their widespread use.
The aim of the present PhD research project is to give a contribution to implement synthetic approaches to develop novel colloidal magnetic nano-architectures with applications in biomedical field. In particular, the research moves along two principal themes. From one side it has been investigated the possibility to tune magnetic
nanoparticles size and properties through appropriate synthetic methodologies. From the other side, since the selected synthetic approach provides hydrophobic
nanoparticles, particular attention has been devoted to the surface modification with organic and inorganic coatings in order to convert them into water dispersible systems.
To this end, a multi-technique approach is used. A first structural analysis is performed by wide-angle X-ray Diffraction to obtain information about the crystalline phases and the crystallite size (coherent domain). Direct images of the samples, obtained by Transmission Electron Microscopy observations allowed an evaluation of the particle sizes, their distribution and the homogeneity of the particle dispersions. In order to study the magnetic-nanoparticles/coating interface, studies with FT-IR spectroscopy have been performed on hydrophobic nanoparticles and on all the systems in which the surface modification has been applied. To investigate the textural
properties of the silica based composites nitrogen physisorption measurements have been performed in combination with low-angle X-ray diffraction to evidence the ordered porosity. Magnetic measurements have been acquired in order to understand how the different chemical composition, size and coatings can affect the magnetic properties of nanoparticles. All the measurements have been carried out at the Department of Chemical and Geological Sciences at the University of Cagliari
Integrated support for Adaptivity and Fault-tolerance in MPSoCs
Get PDFThe technology improvement and the adoption of more and more complex applications in consumer electronics are forcing a rapid increase in the complexity of multiprocessor systems on chip (MPSoCs). Following this trend, MPSoCs are becoming increasingly dynamic
and adaptive, for several reasons. One of these is that applications are getting intrinsically dynamic. Another reason is that the workload on emerging MPSoCs cannot be predicted because modern systems are open to new incoming applications at run-time. A third reason which calls for adaptivity is the decreasing component reliability associated with technology scaling. Components below the 32-nm node are more inclined to temporal or even permanent faults. In case of a malfunctioning system component, the rest of the system is supposed to take over its tasks. Thus, the system adaptivity goal shall influence several de-
sign decisions, that have been listed below: 1) The applications should be specified such that system adaptivity can be easily supported. To this end, we consider Polyhedral Process Networks (PPNs) as model of computation to specify applications. PPNs are composed by concurrent and autonomous processes that communicate between each other using bounded FIFO channels. Moreover, in PPNs the control is completely distributed, as well as the memories. This represents a good match with the emerging MPSoC architectures, in which processing elements and memories are usually distributed. Most importantly, the simple operational semantics of PPNs allows for an easy adoption of system adaptivity mechanisms.
2) The hardware platform should guarantee the flexibility that adaptivity mechanisms require. Networks-on-Chip (NoCs) are emerging communication infrastructures for MPSoCs
that, among many other advantages, allow for system adaptivity. This is because NoCs are generic, since the same platformcan be used to run different applications, or to run the same application with different mapping of processes. However, there is a mismatch between the
generic structure of the NoCs and the semantics of the PPN model. Therefore, in this thesis we investigate and propose several communication approaches to overcome this mismatch.
3) The system must be able to change the process mapping at run-time, using process migration. To this end, a process migration mechanism has been proposed and evaluated.
This mechanism takes into account specific requirements of the embedded domain such as predictability and efficiency. To face the problem of graceful degradation of the system, we enriched the MADNESS NoC platform by adding fault tolerance support at both software and hardware level. The proposed process migration mechanism can be exploited to cope with permanent faults by migrating the processes running on the faulty processing element.
A fast heuristic is used to determine the new mapping of the processes to tiles. The experimental results prove that the overhead in terms of execution time, due to the execution time of the remapping heuristic, together with the actual process migration, is almost negligible
compared to the execution time of the whole application. This means that the proposed approach allows the system to change its performance metrics and to react to faults without a substantial impact on the user experience
Patologie dell’atto amministrativo digitale: profili processuali e prospettive
Get PDFThis research deals with computerised administrative procedures. Particular attention is given to how specific issues such as the adoption of computerized administrative
documents and/or, more generally, the adoption of Information and Communication Technologies (ICT) for administrative activities may affect the validity of computerized administrative documents by the Administration Judge. This study briefly examines the complex and multifaceted legal and regulatory framework concerning administrative procedures and digital administration, Law no. 241/1990 as amended in the following years, the Code of Digital Administration (Law Decree no. 82/2005) and the recently introduced legal provisions which have considerably encouraged the long-awaited creation of a “Digital Italy” (Law Decree no. 5/2012; Law Decree no. 83/2012 and Law Decree no. 179/2012). This paper analyzes the fundamental models for the implementation of administrative actions (ex parte and ex officio applications) and the specific features of their different
steps (start of procedure, taking of evidence, participation of an individual to the electronic procedure, decision-making stage) in light of the use of information
technologies. Particular attention is paid to the decision-making stage of an electronic proceeding, also to the purpose of investigating defects and jurisdictional validity of an electronic proceeding. In this respect the two methods for using the ICT were briefly examined, the
one whereby the ICT role concerns the many operations and formalities aimed at the adoption of the final document, subject to the officer’s will to accept the administrative
decision, and the one whereby the final document is adopted by means of a software program. The limits of the possibility to elaborate the content of administrative
documents are also examined. Illegality aspects of electronic administrative proceedings and pathologies of
automated administrative documents are specially investigated. Special attention is paid
to formal defects and to those defects concerning the signature of electronic administrative documents: signature of an electronic document printed on paper; lack
and disavowal of a signature. This research also examines the possibility to apply traditional legal defects (lack of jurisdiction, abuse of power, infringement of the law) to
the specific area of computerized administrative documents
Il dissesto finanziario negli enti locali alla vigilia dei Decreti Legislativi n. 149/2010 e n. 174/2013. Analisi del fenomeno italiano
Get PDFThis work tries to analyse the phenomenon of financial distress in Italian public local authorities in the light of the entry into force in 1989 of the Italian legislation on the Financial Distress of Public Authorities and of the reform of Title V of the Constitution in 2001, which have produced important changes. After nearly twenty-five years, we wonder what it is the real coercive power of the law on pushing bodies to declare bankruptcy and whether it is possible to chart characteristic features of a possible
profile-type of local authority in default. Finally, with some interviews to the main actors of the public bodies we intend to investigate the different points of view and way
of approaching a financial distress at the time of its official declaration. The methodological approach was exploratory-descriptive, based on the analysis of archival
records and two non-structured interviews with key informants
Ion specific effects on charged interfaces
Get PDFThe physico-chemical phenomena occurring at charged interfaces are specifically affected by the type and the concentration of electrolytes. This has implications both in living and in inorganic systems. The discovery of the ‘ion specific effects’ dates back to Hofmeister (1888), who observed the specific effect of salts in promoting egg white proteins precipitation. Nowadays we are aware that ion specific effects are ubiquitous in all fields of science and technology where electrolytes play a role.
Until few years ago the occurrence of ion specific (Hofmeister) effects, although well recognized in a huge number of cases (i.e. protein precipitation and crystallization, viscosity of aqueous solutions, colloid stability, enzyme activities, etc.), was mainly related to ion-induced change of ‘water structure’. To depict these peculiar behaviors the words ‘kosmotropic’ (order maker) and ‘chaotropic’ (disorder maker) ion were coined. Recently, Ninham and co-workers have proposed an explanation of the Hofmeister phenomena based on the consideration that non- electrostatic ion-specific dispersion forces have to be evaluated at the same level of electrostatics in the Poisson-Boltzmann description of ion-ion and ion-surface interactions. The investigation of the effects that the type of salt and its concentration have on the properties (surface charge and potential, electrochemical behavior) of some proteins (BSA, cytochrome c, lysozyme) and of silica-based ordered mesoporours material (OMM) chosen as model systems, was the main purpose of this thesis. In particular, the present work is devoted to reach the following aims: I) To investigate the ion specific effects due to added electrolytes in different systems where soft or hard matter charged interfaces occur; II) To ascertain the interpretation of different experimental results through Ninham’s theory modeling; III) To show how the comprehension of Hofmeister effects may be used for technological purposes, for instance in the development of ‘controlled release systems’. This thesis can be outlined in two parts. The first part is dedicated to a description of the intermolecular forces in terms of the classical and new theories, which permits to understand ion specific effects at charged interfaces, and to the presentation of the experimental techniques that were used to investigate the various systems. In particular, Chapter 1 gives a brief summary of the classical theories of ‘electrical double layer’ and of ‘colloid stability’ used to describe the phenomena occurring at charged interfaces. Chapter 2 gives an overview of the most important ion-specific phenomena since the famous Hofmeister’s experiment, and of the most recent theoretical approaches currently used to explain ion specificity. Chapter 3 describes the experimental methods and the techniques that were used for the investigation of ion specific effects along the PhD work. In the second part of the thesis the papers where the properties of the charged interfaces were studied as a function of the ionic strength and the type of electrolyte are reported. Papers I and II deal with the model protein Bovine Serum Albumin (BSA) in aqueous solution where a liquid-liquid charged interface is formed.
In paper I the surface charge (Zp) and the Zeta potential (ζ) of BSA were determined by means of potentiometric titrations (PT) and electrophoretic light scattering (ELS), respectively. The measurements were carried on as a function of the pH and the ionic strength of the background electrolyte (NaCl) used in the measurements. The isoionic point (IIP) and the isoelectric point (IEP) of the protein were determined as the intercept of the surface charge and zeta potential curves with pH axis, respectively. The shift in the values of IIP (to higher pH) and of IEP (to lower pH), that occurs with increasing ionic strength, depends on the specific interaction of the ions of the background electrolyte with the surface of BSA. The experimental values of IIP and IEP were compared with the respective theoretical values calculated by means of the charge regulation model. In this model the H+ concentration at the protein surface (surface pH), is calculated introducing a term that takes into account the additional potential due to the dispersion forces proposed by Ninham. The same experimental techniques were used in paper II for the determination of the surface charge (Zp) and the electrophoretic mobility (μe) of BSA in the presence of different sodium salts at fixed ionic strength. In this case the values of electrophoretic mobility were used to calculate the effective charge (Zeff) of the protein by means of Henry’s equation. Zp is the net (proton) charge of the protein in the absence of other bound ions. Zeff, instead, is the charge of the protein including the ions (cations and anions) adsorbed on its surface. The difference between these two values (Zp-Zeff) calculated at a given pH, gives an estimation of ion-binding, that is the number of ions (cations and anions) bound to the protein surface. Paper III focuses on a silica-based Ordered Mesoporous Material (OMM), namely SBA-15, dispersed in an aqueous solution and thus forming a solid-liquid charged interface. Potentiometric titrations (PT) were used to estimate the surface charge density (σ) of silica surface of SBA-15 particles in water at different pH values. The surface charge density/pH curve was studied in the presence of salts with different cations but same anion. The experimental results were compared with the theoretical surface charge density (σ) vs pH curves calculated through Ninham’s theory. The theoretical curves for SBA-15 surface charge density were obtained using Ninham’s charge regulation model mentioned above for Paper I. Paper IV highlights the electrochemical behavior of cytochrome c (cyt c) as a function of salt type and concentration.
The electrochemical properties of cytochrome c, a redox protein, were investigated by means of differential pulse voltammetry (DPV) as a function of the ionic strength and the type of the supporting electrolyte used to guarantee the electrical conductivity of the solution. This work was carried out during my stay at the laboratory of Prof. Edmond Magner at the University of Limerick (Ireland). The redox potential (E°’) and the intensity of the current measured in the electrochemical experiments were firstly investigated as a function of the ionic strength using NaCl as the supporting electrolyte. The values of the redox potential were fitted using a model derived from the Debye-Hückel extended law. Then, the redox potential and the intensity of current were investigated in the presence of different electrolytes (cations and anions of the Hofmeister series) at a fixed ionic strength. The values of redox potential measured at different temperatures were used to obtain thermodynamic information (H°’ and S°’) about the redox process. Paper V reports a possible application of the electrolyte effects on charged interfaces for the development of a “controlled drug release system”. The adsorption and the release of Lysozyme (Lyz), a protein with antimicrobial activity, from two OMMs (SBA-15 and MSE) were investigated. The work focused on the effect of two weak electrolytes (buffers) used to fix the pH of the adsorbing solution. SBA-15 and MSE have both a hexagonal mesoporous structure, but different chemical composition and surface properties. In particular MSE is characterized by a higher “hydrophobic” character due to the presence of bridged ethylene groups linked to the silicon skeleton. The adsorption of the lysozyme on the two OMMs was carried out at two different pHs (pH 7.0 and 9.6) using two suitable buffers. Lysozyme release in ‘in-vitro’ physiological conditions was investigated. It was found that the Lysozyme loading changes very much, depending on the surface charge and on the type of surface-protein interactions. Non-electrostatic interactions, mainly involved in the adsorption/release process at the Lyz/MSE interface, seems to produce a stronger protein binding while electrostatic forces dominate at the Lyz/SBA-15 interface. Another very significant parameter that affects both adsorption and release is the pH of the adsorbing solution. The changing in the charge of the surfaces allows for a potential modulation of the global performance, both in terms of protein’s loaded amount, and in terms of release’s rate. The main conclusions of this work can be sketched by the following important remarks: I) All kinds of soft or hard matter charged interfaces (liquid-liquid and solid-liquid) are strongly affected by ion specific effects, in striking agreement with Hofmeister series (almost always). II) Theoretical modeling based on the modified Poisson-Boltzmann equation that includes ion dispersion potentials calculated introducing ab-initio ion polarizabilities of differently hydrated ions allowed to reproduce experimental electrochemical properties at a semi-quantitative level, thus validating further Ninham’s theoretical approach for intermolecular forces. III) In the case of OMMs used for immobilizing proteins, both adsorption and release can be modulated by the choice of different buffers (that modifies the pH). Ultimately, it can be asserted that the choice of background salt used in the adsorbing solutions may address different biotechnological purposes for instance a drug delivery system or a biocatalyst