Spatial prediction of hourly rainfall via radar calibration is addressed. The change of support problem (COSP), arising when the spatial supports of different data sources do not coincide, is faced in a non-Gaussian setting; in fact, hourly rainfall in Emilia-Romagna region, in Italy, is characterized by abundance of zero values and right-skeweness of the distribution of positive amounts. Rain gauge direct measurements on sparsely distributed locations and hourly cumulated radar grids are provided by the ARPA-SIMC Emilia-Romagna. We propose a three-stage Bayesian hierarchical model for radar calibration, exploiting rain gauges as reference measure. Rain probability and amounts are modeled via linear relationships with radar in the log scale; spatial correlated Gaussian effects capture the residual information. We employ a probit link for rainfall probability and Gamma distribution for rainfall positive amounts; the two steps are joined via a two-part semicontinuous model. Three model specifications differently addressing COSP are presented; in particular, a stochastic weighting of all radar pixels, driven by a latent Gaussian process defined on the grid, is employed. Estimation is performed via MCMC procedures implemented in C, linked to R software. Communication and evaluation of probabilistic, point and interval predictions is investigated. A non-randomized PIT histogram is proposed for correctly assessing calibration and coverage of two-part semicontinuous models. Predictions obtained with the different model specifications are evaluated via graphical tools (Reliability Plot, Sharpness Histogram, PIT Histogram, Brier Score Plot and Quantile Decomposition Plot), proper scoring rules (Brier Score, Continuous Rank Probability Score) and consistent scoring functions (Root Mean Square Error and Mean Absolute Error addressing the predictive mean and median, respectively). Calibration is reached and the inclusion of neighbouring information slightly improves predictions. All specifications outperform a benchmark model with incorrelated effects, confirming the relevance of spatial correlation for modeling rainfall probability and accumulation.

This thesis focuses on the analysis of heavy-tailed distributions, which are widely applied to model phenomena in many disciplines. The definition of heavy tails based on the theory of regular variation highlights the importance of the tail index, which indicates the existence of moments and characterises the rate at which the tail decays. Two new approaches to make inference for the tail index are proposed. The first approach employs a regression technique and constructs an estimator of the tail index. It exploits the fact that the behaviour of the characteristic function near the origin reflects the behaviour of the distribution function at infinity. The main advantage of this approach is that it utilises all observations to constitute each point in the regression, not just extreme values. Moreover, the approach does not rely on prior information on the starting point of the tail behaviour of the underlying distribution and shows excellent performance in a wide range of cases: Pareto distributions, heavy-tailed distributions with a non-constant slowly varying factor, and composite distributions with heavy tails. The second approach is motivated by the asymptotic properties of a special moment statistic, the so-called partition function. This statistic considers blocks of data and is generally used in the context of multifractality. Due to the interplay between the weak law of large numbers and the generalised central limit theorem, the asymptotic behaviour of the partition function is strongly affected by the existence of moments even for weakly dependent samples. Via a quantity, the scaling function, a graphical method to identify the existence of heavy tails is proposed. Moreover, the plot of the scaling function allows one to make inference for the underlying distribution: with infinite variance, finite variance with tail index larger than two, or all moments finite. Furthermore, since the tail index is reflected at the breakpoint of the plot of the scaling function, this gives the possibility to estimate the tail index. Both these two approaches use the entire distribution, not just the tail, to analyse the tail behaviour. This sheds a new light on the analysis of heavy-tailed distributions. At the end of this thesis, these two approaches are used to detect power laws in empirical data sets from a variety of fields and contribute to the debate on whether city sizes are better approximated by a power law or a log-normal distribution.

In questa tesi, l’analisi verte su risk-adjusted performance, proprietà statistiche e caratteristiche dei fondi hedge (FH). Nel primo articolo, i risultati relativi al survivorship bias e backfill bias indicano che l’impatto delle distorsioni è diverso a seconda delle strategie. Utilizzando il modello multifattoriale di Fung and Hsieh (2004), l’analisi della performance indica che il 42% dei FH ha ottenuto un rendimento superiore al mercato. Infine, utilizzando dei metodi parametrici e non parametrici, l’analisi della persistenza indica differenti livelli di persistenza a seconda della strategia. Nel secondo articolo, vengono analizzati i fondi di fondi hedge (FOHFs). I risultati sono particolarmente interessanti. In primo luogo, i FOHFs e le sotto strategie hanno generato un excess return positivo; inoltre l’alfa ottenuto attraverso il modello a 7 fattori di Fung and Hsieh (2004) risulta elevato. In secondo luogo, i FOHFs e le sotto strategie hanno un rendimento inferiore a quello dell’indice dei FH. In terzo luogo, le correlazioni tra gli indici dei FOHFs e l’indice azionario sono inferiori rispetto alle correlazioni tra l’indice dei FH e gli indici azionari. Infine, l’indice dei FH e quelli dei FOHFs sono positivamente correlati con l’indice azionario quando il mercato tende al ribasso, ma risultano non correlati con l’indice azionario quando il mercato tende al rialzo. Rispetto all’indice dei FH, gli indici dei FOHFs hanno una correlazione minore con gli indici azionari in entrambe le fasi del mercato, suggerendo che i FOHFs forniscono benefici maggiori in termini di diversificazione rispetto ai fondi hedge puri. / In this thesis, I examine the risk-adjusted performance, statistical properties and fund characteristics of hedge fund investments. In Essay One, results of survivorship bias and backfill bias by investment styles indicate that biases are different across styles. Using a multi-factor model of Fung and Hsieh (2004), the analysis of performance indicates that 42% of the hedge funds significantly outperformed the market. Finally, using parametric and non-parametric methods, the analysis of persistence indicates different degree of persistence depending on the hedge fund strategy. In Essay Two, I analyse fund of hedge funds (FOHFs). I find several interesting results. First, FOHFs and the sub-strategies earn positive excess returns and a high Fung and Hsieh 7-factor alpha. Second, FOHFs and the sub-strategies underperform the hedge fund index (HFI). Third, the correlations between FOHF indices and equity index are lower than correlations between HFI and equity indices. Finally, hedge funds and FOHFs are positively correlated with the equity index in the bear markets but uncorrelated with the equity index in the bull markets. Compared to HFI, FOHF indices have lower correlation with equity index in both bull and bear markets, indicating that FOHFs provide better diversification benefits than individual hedge funds.

Smart energy city (SEC) development is a component of the urban development initiative smart city, which has been a popular response to the global energy challenge in Europe during the past two decades. SEC development aims to increase the sustainability of urban energy systems and services. Since 2011, SEC development has been supported by the European Commission as part of the Strategic Energy Technology plan (SET-Plan) and through the European Union Programmes for Research and Technological Development (specifically FP7 and Horizon 2020). This, along with the promising vision of SEC development and considerable financial support by the private sector, has encouraged numerous European cities to initiate SEC projects. Successful implementation of these projects at the urban scale is crucial to achievement of urban energy objectives and sustainability of future urban development. The here presented thesis aims to support urban decision-makers towards successful implementation of urban scale smart energy city development in Europe. The study includes three stages. The first stage is dedicated to conceptual analysis. Within this stage, I conceptualized smart city through a keyword analysis of existing literature on the concept. Then, within the context of the smart city concept, I defined SEC development through literature review and expert knowledge elicitation. The second stage is dedicated to empirical investigation. Using the definition of SEC development, I distinguished and investigated 43 previously implemented SEC projects to identify common barriers that hinder successful implementation of SEC development. In addition, I proposed a new multi-dimensional methodology that allows a simultaneous prioritization of barriers against their probability, the level of impact, scale, origin, and relationship with other barriers. The third stage of the thesis is dedicated to learning methodologies that allow efficient transfer of knowledge from the past SEC experiences to the new SEC developments. I introduced the application of two learning methodologies that support decision-makers to predict barriers to the implementation of a new SEC project: case-based learning and decision tree learning. The former predicts barriers based on internal similarities between the new SEC project and the past projects. The latter uses the past projects and creates a predictive model for each barrier based on internal and external project characteristics. These models are later used to predict barriers to a new SEC project. Both methodologies were tested in a new SEC project, named SINFONIA. The conceptual analysis revealed that application of information and communication technologies, the collaboration of multiple stakeholders, integration of multiple urban domains, and sustainability evaluation are the constant characteristics (i.e. principles) of smart city and SEC development. It resulted in, to the best of my knowledge, the first multi-dimensional and comprehensive definition of SEC development, revealing its principles, objectives, domains of intervention, stakeholders, time and spatial dimensions. Furthermore, a list of smart energy solutions in each SEC domain of intervention was provided. The empirical investigation of the past SEC projects resulted in the identification of 35 common barriers to the implementation of SEC development, categorized in policy, administrative, legal, financial, market, environmental, technical, social, and information and awareness dimensions. The barrier prioritization showed that barriers related to collaborative planning, external funding of the project, providing skilled personnel, and fragmented ownership should be the key action priorities for SEC project coordinators. Application of case-based learning methodology resulted in identifying five past SEC projects that were the most similar to the SINFONIA project in terms of project internal characteristics. Investigating the barriers to the similar projects revealed that fragmented ownership is the most probable barrier to implementation of SINFONIA project. Application of the decision trees methodology resulted in generation of 20 barrier models, four of which showed a very good performance in prediction of barriers: lack of values and interest in energy optimization measures, time-consuming requirements by European Commission concerning reporting and accountancy, economic crisis, and local unfavorable regulations for innovative technologies. None of these four barriers were predicted to occur in the SINFONIA project. The application of this method in the SINFONIA showed a higher predicting power when a barrier was absent. The findings of the here presented thesis contribute to successful implementation of SEC development by supporting decision-makers in different phases of SEC projects. The results of the conceptual analysis contribute to a common understanding and foster the dialogue on the concept among various SEC stakeholders, particularly decision-makers and urban planners. The results of the empirical investigation lead to a better comprehension and evaluation of the barriers to the implementation of SEC projects in order to efficiently allocate resources to mitigate barriers. The proposed learning methodologies proved to be promising in helping decision-makers to identify similar projects to a new SEC development and to predict barriers to the implementation of new SEC projects. The thesis concludes that SEC is an outstanding urban development that can make a valuable contribution to the sustainability of urban energy systems. The specific characteristics of SEC development pose new challenges to the future smart and sustainable urban planning. Nevertheless, SEC development brings about unprecedented opportunities for integration and application of advanced quantitative techniques with current urban planning methods. This allows efficient knowledge transfer in not only intra-urban but also inter-urban levels in order to provide a collaborative, integrated and constructive movement towards successful implementation of SEC projects and sustainability of future urban development.