Probabilistic Network Analysis Research Articles

Probabilistic network analysis of social-ecological relationships emerging from EU LIFE projects for nature and biodiversity: An application of ERGM models in the case study of the Veneto region (Italy)

Considering social-ecological relationships in managing protected areas is fundamental to ensuring effective biodiversity conservation and restoration governance. Network analysis offers valuable methods to disentangle intangible relations between and within the social and ecological systems. In this way, it could be possible to identify and integrate multiple social and ecological variables that inevitably affect collaborative environmental governance's effectiveness. Nevertheless, this research area is still nascent, with few methodologies and concrete applications reported in the scientific literature. With this study, we aim to propose a robust novel application of a network methodology to enrich the evaluation of the effectiveness of collaborative environmental governance for nature and biodiversity, which has been applied through the analysis of social-ecological relationships that emerged from EU-cofounded LIFE-NAT projects. Specifically, we focus on LIFE-NAT projects implemented in the Veneto Region (Italy) financed in the last programming period (2014–2020). Through formulating four research hypotheses to be tested through Exponential Random Graph Models, we analyze 13 LIFE-NAT projects involving 83 social actors and 29 Natura 2000 (N2000) sites composed of 57 protected habitats. Results show that LIFE-NAT projects in Veneto Region stimulate polycentric governance. Nevertheless, they still need to concretize a multi-actor and multilevel governance. Furthermore, the analysis highlights that selected LIFE-NAT projects implement activities in N2000 sites able to support ecological connectivity and synergies across marine, freshwater, and land habitats through the bridging role of forests, especially in estuarine and coastal areas.

Bayesian analysis of cytokines and chemokine identifies immune pathways of HBsAg loss during chronic hepatitis B treatment

Our objective was to examine differences in cytokine/chemokine response in chronic hepatitis B(CHB) patients to understand the immune mechanism of HBsAg loss (functional cure) during antiviral therapy. We used an unbiased machine learning strategy to unravel the immune pathways in CHB nucleo(t)side analogue-treated patients who achieved HBsAg loss with peg-interferon-α(peg-IFN-α) add-on or switch treatment in a randomised clinical trial. Cytokines/chemokines from plasma were compared between those with/without HBsAg loss, at baseline, before and after HBsAg loss. Peg-IFN-α treatment resulted in higher levels of IL-27, IL-12p70, IL-18, IL-13, IL-4, IL-22 and GM-CSF prior to HBsAg loss. Probabilistic network analysis of cytokines, chemokines and soluble factors suggested a dynamic dendritic cell driven NK and T cell immune response associated with HBsAg loss. Bayesian network analysis showed a dominant myeloid-driven type 1 inflammatory response with a MIG and I-TAC central module contributing to HBsAg loss in the add-on arm. In the switch arm, HBsAg loss was associated with a T cell activation module exemplified by high levels of CD40L suggesting T cell activation. Our findings show that more than one immune pathway to HBsAg loss was found with peg-IFN-α therapy; by myeloid-driven Type 1 response in one instance, and T cell activation in the other.

A network analytical framework to analyze infrastructure damage based on earthquake cascades: A study of earthquake cases in Japan

Earthquakes can result in disasters affecting infrastructure and assets that are exposed and vulnerable and thus prone to damage. Considering earthquake cascades, we can reveal the immediate and intermediate effects of earthquakes causing infrastructure damage. In this study, we propose a network analytical modeling approach to analyze earthquake cascades. The analytical framework integrates incident causation modeling and probabilistic network analysis. The linear effects from earthquakes are delineated in incident chains to represent the consequences of earthquakes to infrastructure; the incident chains are then remodeled topologically by the network theory to reveal the nonlinear cascading effects of earthquakes to infrastructure. The network of earthquakes’ effects and intermediate effects is then structurally analyzed by probabilistic network analysis that uncovers the unifying principles and statistical topology of earthquake cascades to infrastructure. The results show that earthquakes can directly cause significant damage to infrastructure and indirectly cascade the effects to infrastructure damage via intermediate effects; each effect or intermediate effects in earthquake cascades have their own role in leading to infrastructure damage. The critical infrastructure damage due to earthquakes and critical intermediate effects in propagating earthquake cascades are revealed in this study by the network analysis of network density, degree centrality, closeness centrality, betweenness centrality, and cascade quantification. To reduce infrastructure damage due to earthquake cascades, we suggest tackling the directed interdependent intermediate effects to minimize the risks associated with earthquakes to infrastructure damage. The network analytical framework in this study can be used to understand the propagation damage of earthquakes on infrastructure; thus, our work can provide valuable insight into earthquake preparedness and emergency management.

Matrix Estimation Meets Statistical Network Analysis: Extracting low-dimensional structures in high dimension

The study of complex relationships among the elements of a large collection of random variables lead to the development of a number of areas in probability and statistics such as probabilistic network analysis or random matrix theory. The aim of the workshop was to address the challenge to develop a coherent mathematical framework within which these areas can be integrated, for a successful analysis of massive and complicated data sets.

Probabilistic convergence guarantees for type-II pulse-coupled oscillators

We show that a large class of pulse-coupled oscillators converge with high probability from random initial conditions on a large class of graphs with time delays. Our analysis combines previous local convergence results, probabilistic network analysis, and a classification scheme for type-II phase response curves to produce rigorous lower bounds for convergence probabilities based on network density. These results suggest methods for the analysis of pulse-coupled oscillators, and provide insights into the balance of excitation and inhibition in the operation of biological type-II phase response curves and also the design of decentralized and minimal clock synchronization schemes in sensor nets.

A proposal on improved procedures for estimating task-time distributions in PERT

The objective of using the well-known PERT task-time estimation formulas has two components: (i) to elicit subjective time estimates from an ‘expert’; and (ii) to translate these estimates into ‘something’ that can be used as inputs to probabilistic network analysis. By incorporating the more current techniques in subjective probability elicitation and distribution theory, this paper proposes and justifies a practical and more logical two-step procedure for achieving both components of the objective. The first step uses a seven-point fractile-estimation procedure, the second step determines the parameters of a beta and a Ramberg-Schmeiser distribution function that minimize the mean square deviation between the distribution function and the estimated fractiles. The procedure can be easily implemented by practitioners using a canned program.

Probabilistic network analysis for project completion time∗

Abstract The paper deals with the probabilistic network analysis for project completion time for various networks. As the project completion time depends on the path with the longest total completion time of all the given paths, it is expressed as a function of various activity durations according to known activity precedence relationship in the present study. As all the activity durations are considered probabilistic in nature, the total completion time is also a probabilistic variable and the determination of that requires rigorous mathematical analysis. Rosenblueth's two point estimates scheme is used for this non-linear probabilistic analysis problem and the results are compared with Monte Carlo Simulation. Such a study is conducted for two different networks and the results compared very well for both sample problems, both for uniform and normal distribution of the random variables which are the activity durations and the activity precedence relationships in the present study.