Search

Bayesian networks are widely used to learn and reason about the dependence structure of discrete variables. However, they can only formally encode symmetric conditional independence, which is often too strict to hold in practice. Asymmetry-labeled DAGs have been recently proposed to extend the class of Bayesian networks by relaxing the symmetric assumption of independence and denoting the dependence between the variables of interest. Here, we introduce novel structural learning algorithms for this class of models, which, whilst efficient, allow for a straightforward interpretation of the underlying dependence structure. A comprehensive computational study highlights the efficiency of the algorithms. A real-world data application using data from the Fear of COVID-19 Scale collected in Italy showcases their use in practice. [ABSTRACT FROM AUTHOR]

Due to the dramatic health situation caused by the COVID-19 pandemic, in Italy the emergency remote teaching lasted longer than in other countries. The mandatory teaching modalities have required digital transformation processes in a framework where digital-divide is one of the limitations to school modernization. We believe that the experience can promote a deeper formatting of organizational process. The paper shows results of a multitarget research carried out during the Italian lockdown aiming at animating the debate around school from multi-actors perspectives and at supporting policies. The paper aims at showing the potentiality of a multivariate statistical method as a tool supporting school managers in identifying those challenges they have to face to improve the setting up of internal processes. The main result is a model supporting the decision making process at orienting school managers strategies. [ABSTRACT FROM AUTHOR]

The measure of the equitable and sustainable well-being (Bes) is of growing interest in the last years. The National Institute of Statistics (Istat) provides, for Italy, a wide set of indicators describing each domain of well-being that is, by definition, a multidimensional concept. In this study, we propose the use of Bayesian networks to deal with basic and composite Bes indicators. Its capability to model very complex multivariate dependence structures is useful to describe the relationships between indicators belonging to different domains and, being a probabilistic expert system, the estimated network could be also useful for probabilistic inference and what-if analysis. In this study, all the Bayesian networks structures have been estimated by means of the hill climbing algorithm based on bootstrap resampling and model averaging in order to prevent bias due to deviations from the normality assumption. [ABSTRACT FROM AUTHOR]

In this paper, a structure learning algorithm is applied to the sensory analysis field to study the factors that have an influence on the quality of Italian wines. Directed acyclic graphs, involving chemical as well as sensory variables, will be proposed to suggest hypotheses about causal connections between these variables and the Altroconsumo’s Global Score of Quality, given by the Italian independent consumer’s association Altroconsumo in its annual publication Guida Vini (Wines’ Guide). The analysis is performed considering all types of wine included in the database, as well as red and white wines separately. [ABSTRACT FROM AUTHOR]

Abstract: The maintenance of organic farming production schemes is a theme receiving a growing interest now that there are signs of a slowing in organic farming uptake in Italy. The present study develops a model based on a Bayesian network (BN) that is aimed at investigating the factors that affect the exit of a farm from the organic sector and to simulate the probability of maintaining an organic scheme for different farm types. The model is based on a database of organic farms, which has been integrated with qualitative information. Farm-type simulation and sensitivity analysis of most of the relevant variables have been carried out. Main results show that arable farm types are those with a higher probability to stay in the organic sector, while farmers’ age, Province the farm is situated in and farm size are the factors mostly influencing probability scores. [Copyright &y& Elsevier]

The management of Operational Risk has been a difficult task due to the lack of data and the high number of variables. In this project, we treat operational risks as multivariate variables. In order to model them, copula functions are employed, which are a widely used tool in finance and engineering for building flexible joint distributions. The purpose of this research is to propose a new methodology for modelling Operational Risks and estimating the required capital. It combines the use of graphical models and the use of copula functions along with hyper-Markov law. Historical loss data of an Italian bank is used, in order to explore the methodology’s behaviour and its potential benefits. [ABSTRACT FROM AUTHOR]

The response to the COVID-19 pandemic has been highly variable. Governments have applied different mitigation policies with varying effect on social and economic measures, over time. This article presents a methodology for examining the effect of mobility restriction measures and the association between health and population activity data. As case studies, we refer to the pre-vaccination experience in Italy and Israel. Facing the pandemic, Israel and Italy implemented different policy measures and experienced different population behavioral patterns. Data from these countries are used to demonstrate the proposed methodology. The analysis we introduce in this paper is a staged approach using Bayesian Networks and Structural Equations Models. The goal is to assess the impact of pandemic management and mitigation policies on pandemic spread and population activity. The proposed methodology models data from health registries and Google mobility data and then shows how decision makers can conduct scenario analyses to help design adequate pandemic management policies.

In this paper, we focus on a Bayesian network s approach to combine traditional survey and social network data and official statistics to evaluate well-being. Bayesian networks permit the use of data with different geographical levels (provincial and regional) and time frequencies (daily, quarterly, and annual). The aim of this study was twofold: to describe the relationship between survey and social network data and to investigate the link between social network data and official statistics. Particularly, we focused on whether the big data anticipate the information provided by the official statistics. The applications, referring to Italy from 2012 to 2017, were performed using ISTAT's survey data, some variables related to the considered time period or geographical levels, a composite index of well-being obtained by Twitter data, and official statistics that summarize the labor market.