Dynamics of coastal aquifers: data-driven forecasting and risk analysis

This work is focused on the study of saltwater intrusion in coastal aquifers, and in particular on the realization of conceptual schemes to evaluate the risk associated with it. Saltwater intrusion depends on different natural and anthropic factors, both presenting a strong aleatory behaviour, that should be considered for an optimal management of the territory and water resources. Given the uncertainty of problem parameters, the risk associated with salinization needs to be cast in a probabilistic framework. On the basis of a widely adopted sharp interface formulation, key hydrogeological problem parameters are modeled as random variables, and global sensitivity analysis is used to determine their influence on the position of saltwater interface. The analyses presented in this work rely on an efficient model reduction technique, based on Polynomial Chaos Expansion, able to combine the best description of the model without great computational burden. When the assumptions of classical analytical models are not respected, and this occurs several times in the applications to real cases of study, as in the area analyzed in the present work, one can adopt data-driven techniques, based on the analysis of the data characterizing the system under study. It follows that a model can be defined on the basis of connections between the system state variables, with only a limited number of assumptions about the "physical" behaviour of the system.

Modelling and Uncertainty Quantification application to SA simulation codes in advanced SMR

In the framework of a global transition to a low-carbon energy mix, the interest in advanced nuclear Small Modular Reactors (SMRs) has been growing at the international level. Due to the high level of maturity reached by Severe Accident Codes for currently operating rectors, their applicability to advanced SMRs is starting to be studied. Within the present work of thesis and in the framework of a collaboration between ENEA, UNIBO and IRSN, an ASTEC code model of a generic IRIS reactor has been developed. The simulation of a DBA sequence involving the operation of all the passive safety systems of the generic IRIS has been carried out to investigate the code model capability in the prediction of the thermal-hydraulics characterizing an integral SMR adopting a passive mitigation strategy. The following simulation of 4 BDBAs sequences explores the applicability of Severe Accident Codes to advance SMRs in beyond-design and core-degradation conditions. The uncertainty affecting a code simulation can be estimated by using the method of Input Uncertainty Propagation, whose application has been realized through the RAVEN-ASTEC coupling and implementation on an HPC platform. This probabilistic methodology has been employed in a study of the uncertainty affecting the passive safety system operation in the DBA simulation of ASTEC, providing a further characterization of the thermal-hydraulics of this sequence. The application of the Uncertainty Quantification method to early core-melt phenomena has been investigated in the framework of a BEPU analysis of the ASTEC simulation of the QUENCH test-6 experiment. A possible solution to the encountered challenges has been proposed through the application of a Limit Surface search algorithm.

Machine learning tools for protein annotation: the cases of transmembrane β-barrel and myristoylated proteins

Biology is now a “Big Data Science” thanks to technological advancements allowing the characterization of the whole macromolecular content of a cell or a collection of cells. This opens interesting perspectives, but only a small portion of this data may be experimentally characterized. From this derives the demand of accurate and efficient computational tools for automatic annotation of biological molecules. This is even more true when dealing with membrane proteins, on which my research project is focused leading to the development of two machine learning-based methods: BetAware-Deep and SVMyr. BetAware-Deep is a tool for the detection and topology prediction of transmembrane beta-barrel proteins found in Gram-negative bacteria. These proteins are involved in many biological processes and primary candidates as drug targets. BetAware-Deep exploits the combination of a deep learning framework (bidirectional long short-term memory) and a probabilistic graphical model (grammatical-restrained hidden conditional random field). Moreover, it introduced a modified formulation of the hydrophobic moment, designed to include the evolutionary information. BetAware-Deep outperformed all the available methods in topology prediction and reported high scores in the detection task. Glycine myristoylation in Eukaryotes is the binding of a myristic acid on an N-terminal glycine. SVMyr is a fast method based on support vector machines designed to predict this modification in dataset of proteomic scale. It uses as input octapeptides and exploits computational scores derived from experimental examples and mean physicochemical features. SVMyr outperformed all the available methods for co-translational myristoylation prediction. In addition, it allows (as a unique feature) the prediction of post-translational myristoylation. Both the tools here described are designed having in mind best practices for the development of machine learning-based tools outlined by the bioinformatics community. Moreover, they are made available via user-friendly web servers. All this make them valuable tools for filling the gap between sequential and annotated data.

Digital Forensics AI: on Practicality, Optimality, and Interpretability of Digital Evidence Mining Techniques

Digital forensics as a field has progressed alongside technological advancements over the years, just as digital devices have gotten more robust and sophisticated. However, criminals and attackers have devised means for exploiting the vulnerabilities or sophistication of these devices to carry out malicious activities in unprecedented ways. Their belief is that electronic crimes can be committed without identities being revealed or trails being established. Several applications of artificial intelligence (AI) have demonstrated interesting and promising solutions to seemingly intractable societal challenges. This thesis aims to advance the concept of applying AI techniques in digital forensic investigation. Our approach involves experimenting with a complex case scenario in which suspects corresponded by e-mail and deleted, suspiciously, certain communications, presumably to conceal evidence. The purpose is to demonstrate the efficacy of Artificial Neural Networks (ANN) in learning and detecting communication patterns over time, and then predicting the possibility of missing communication(s) along with potential topics of discussion. To do this, we developed a novel approach and included other existing models. The accuracy of our results is evaluated, and their performance on previously unseen data is measured. Second, we proposed conceptualizing the term “Digital Forensics AI” (DFAI) to formalize the application of AI in digital forensics. The objective is to highlight the instruments that facilitate the best evidential outcomes and presentation mechanisms that are adaptable to the probabilistic output of AI models. Finally, we enhanced our notion in support of the application of AI in digital forensics by recommending methodologies and approaches for bridging trust gaps through the development of interpretable models that facilitate the admissibility of digital evidence in legal proceedings.

Strength and ductility of corroded strands within the framework of safety assessment of existing bridges

Existing bridges built in the last 50 years face challenges due to states far different than those envisaged when they were designed, due to increased loads, ageing of materials, and poor maintenance. For post-tensioned bridges, the need emerged for reliable engineering tools for the evaluation of their capacity in case of steel corrosion due to lack of mortar injection. This can lead to sudden brittle collapses, highlighting the need for proper maintenance and monitoring. This thesis proposes a peak strength model for corroded strands, introducing a “group coefficient” that aims at considering corrosion variability in the wires constituting the strands. The application of the introduced model in a deterministic approach leads to the proposal of strength curves for corroded strands, which represent useful engineering tools for estimating their maximum strength considering both geometry of the corrosion and steel material parameters. Together with the proposed ultimate displacement curves, constitutive laws of the steel material reduced by the effects of corrosion can be obtained. The effects of corroded strands on post-tensioned beams can be evaluated through the reduced bending moment-curvature diagram accounting for these reduced stress-strain relationships. The application of the introduced model in a probabilistic approach allows to estimate peak strength probability functions and consecutive design-oriented safety factors to consider corrosion effects in safety assessment verifications. Both approaches consider two procedures that are based on the knowledge level of the corrosion in the strands. On the sidelines of this main research line, this thesis also presents a study of a seismic upgrading intervention of a case-study bridge through HDRB isolators providing a simplified procedure for the identification of the correct device. The study also investigates the effects due to the variability of the shear modulus of the rubber material of the HDRB isolators on the structural response of the isolated bridge.

New perspectives in statistical mechanics and high-dimensional inference

The main purpose of this thesis is to go beyond two usual assumptions that accompany theoretical analysis in spin-glasses and inference: the i.i.d. (independently and identically distributed) hypothesis on the noise elements and the finite rank regime. The first one appears since the early birth of spin-glasses. The second one instead concerns the inference viewpoint. Disordered systems and Bayesian inference have a well-established relation, evidenced by their continuous cross-fertilization. The thesis makes use of techniques coming both from the rigorous mathematical machinery of spin-glasses, such as the interpolation scheme, and from Statistical Physics, such as the replica method. The first chapter contains an introduction to the Sherrington-Kirkpatrick and spiked Wigner models. The first is a mean field spin-glass where the couplings are i.i.d. Gaussian random variables. The second instead amounts to establish the information theoretical limits in the reconstruction of a fixed low rank matrix, the “spike”, blurred by additive Gaussian noise. In chapters 2 and 3 the i.i.d. hypothesis on the noise is broken by assuming a noise with inhomogeneous variance profile. In spin-glasses this leads to multi-species models. The inferential counterpart is called spatial coupling. All the previous models are usually studied in the Bayes-optimal setting, where everything is known about the generating process of the data. In chapter 4 instead we study the spiked Wigner model where the prior on the signal to reconstruct is ignored. In chapter 5 we analyze the statistical limits of a spiked Wigner model where the noise is no longer Gaussian, but drawn from a random matrix ensemble, which makes its elements dependent. The thesis ends with chapter 6, where the challenging problem of high-rank probabilistic matrix factorization is tackled. Here we introduce a new procedure called "decimation" and we show that it is theoretically to perform matrix factorization through it.

On the Wick product and Wong-Zakai approximations.

This thesis is a compilation of 6 papers that the author has written together with Alberto Lanconelli (chapters 3, 5 and 8) and Hyun-Jung Kim (ch 7). The logic thread that link all these chapters together is the interest to analyze and approximate the solutions of certain stochastic differential equations using the so called Wick product as the basic tool. In the first chapter we present arguably the most important achievement of this thesis; namely the generalization to multiple dimensions of a Wick-Wong-Zakai approximation theorem proposed by Hu and Oksendal. By exploiting the relationship between the Wick product and the Malliavin derivative we propose an original reduction method which allows us to approximate semi-linear systems of stochastic differential equations of the Itô type. Furthermore in chapter 4 we present a non-trivial extension of the aforementioned results to the case in which the system of stochastic differential equations are driven by a multi-dimensional fraction Brownian motion with Hurst parameter bigger than 1/2. In chapter 5 we employ our approach and present a “short time” approximation for the solution of the Zakai equation from non-linear filtering theory and provide an estimation of the speed of convergence. In chapters 6 and 7 we study some properties of the unique mild solution for the Stochastic Heat Equation driven by spatial white noise of the Wick-Skorohod type. In particular by means of our reduction method we obtain an alternative derivation of the Feynman-Kac representation for the solution, we find its optimal Hölder regularity in time and space and present a Feynman-Kac-type closed form for its spatial derivative. Chapter 8 treats a somewhat different topic; in particular we investigate some probabilistic aspects of the unique global strong solution of a two dimensional system of semi-linear stochastic differential equations describing a predator-prey model perturbed by Gaussian noise.