GRETA PANUNZI

Dottoressa di ricerca

ciclo: XXXVII


supervisore: Giovanna Jona Lasinio

Titolo della tesi: Spatial Point Pattern Estimation Using Integrated Laplace Approximation: Methodology and Applicaions

This thesis presents the primary research projects conducted during my PhD program. All projects focus on the challenges of performing spatial analysis by integrating data from diverse sources, a topic that is receiving increasing attention in statistical research due to the exponential growth of data collection technologies, such as social networks and Citizen Science campaigns. The rising demand for advanced analytical methods is driven by the proliferation of devices capable of recording fine-scale spatial data with high temporal resolution, enabling detailed exploration of phenomena across various fields, including ecology, epidemiology, geology, and economics. The thesis begins with an introduction to spatial statistics, outlining its historical development and core concepts. This section also presents the core principles of Bayesian statistics and outlines the applied context of my research. Chapter 2 explores the spatial analysis of processes, discussing key properties of spatial process analysis with a focus on one of its most widely applied types: spatial point processes. Point process models are a natural choice for species distribution modeling when data are represented as point events (presence-only data). This chapter examines a specific type of point process known as Poisson processes, with a particular focus on Log-Gaussian Cox processes. These processes are widely used in point process modeling because they incorporate the properties of the multivariate normal distribution within the framework of Cox processes. In addition, the chapter addresses the computational challenges inherent in spatial models by analyzing strategies in the Bayesian framework, starting from MCMC-based methods to the latest Integrated Nested Laplace approximation method. Chapter 3 presents three applications of the methodology developed during the research period. These applications are based on the data integration methodology defined by Martino, Pace, et al. (2021), highlighting advancements that enable its use in increasingly diverse contexts. Establishing clear protocols that outline the correct estimates’ methodologies to be followed is crucial for studies related to wildlife conservation. It is essential to utilize modern data sources to address the limitations of traditional protocols. However, non-standard data are difficult to use and are often biased by human activity. Therefore, it is necessary to define accurate estimation procedures that account for all potential sources of bias and produce the most accurate estimates possible. Finally, Chapter 4 summarizes the original contributions of the thesis and emphasizes the evolving role of statistical modeling in addressing modern, datadriven challenges. The thesis underscores the importance of continuously adapting and innovating statistical methodologies in response to technological advancements and the increasing complexity of datasets.

Produzione scientifica

11573/1728995 - 2024 - Species distribution modeling approach for biased citizen science data
Panunzi, Greta; Belmont, Jafet; Illian, Janine; Martino, Sara - 04b Atto di convegno in volume
congresso: Sis 2024 (Bari)
libro: Methodological and Applied Statistics and Demography – SIS 2024 Short Papers - ()

11573/1715632 - 2024 - Estimating the spatial distribution of the white shark in the Mediterranean Sea via an integrated species distribution model accounting for physical barriers
Panunzi, Greta; Moro, Stefano; Marques, Isa; Martino, Sara; Colloca, Francesco; Ferretti, Francesco; Jona Lasinio, Giovanna - 01a Articolo in rivista
rivista: ENVIRONMETRICS (-West Sussex : Wiley-Blackwell - West Sussex : John Wiley & Sons -London [Ont.] : Environmetrics Press, 1990-) pp. - - issn: 1099-095X - wos: WOS:001264354400001 (2) - scopus: (0)

11573/1722537 - 2024 - A Pogit model for underreported counts of violence against women in Italy
Polettini, Silvia; Martino, Sara; Panunzi, Greta - 04b Atto di convegno in volume
congresso: 52nd Scientific Meeting of the Italian Statistical Society (SIS 2024) (Bari, Italy)
libro: Methodological and Applied Statistics and Demography I - SIS 2024, Short Papers, Plenary and Specialized Sessions - (978-3-031-64345-3)

11573/1685556 - 2023 - A species distribution modelling framework for combining citizen science data from different monitoring schemes
Belmont, Jafet; Martino, Sara; Panunzi, Greta; Illian, Janine - 04d Abstract in atti di convegno
congresso: GRASPA2023 (PALERMO)
libro: Proceedings of the GRASPA 2023 Conference - (9791221033892)

11573/1696908 - 2023 - A new species distribution modeling approach for biased citizen science data
Panunzi, Greta; Belmont, Jafet; Illian, Janine; Martino, Sara - 04d Abstract in atti di convegno
congresso: International Conference on Statistics and Data Science (ICSDS) (Lisbon)
libro: 2023 IMS International Conference on Statistics and Data Science (ICSDS) - (978-0-940600-86-7)

11573/1687088 - 2023 - A preliminary study on shape descriptors for the characterization of microplastics ingested by fish
Panunzi, Greta; Valente, Tommaso; Matiddi, Marco; Jona Lasinio, Giovanna - 04b Atto di convegno in volume
congresso: Statistical learning sustainability and impact evaluation 2023 (Ancona)
libro: Seas in 2023. Book of the short papers - (9788891935618)

11573/1665434 - 2023 - Modeling of Hardy-Weinberg Equilibrium Using Dynamic Random Networks in an ABM Framework
Tarantino, Riccardo; Panunzi, Greta; Romano, Valentino - 04b Atto di convegno in volume
congresso: International Conference on Complex Networks and Their Applications (Palermo)
libro: Complex Networks and Their Applications XI - (978-3-031-21130-0)

11573/1654206 - 2022 - Seasonal distribution of an opportunistic apex predator (Tursiops truncatus) in marine coastal habitats of the Western Mediterranean Sea
Pace, Daniela Silvia; Panunzi, Greta; Arcangeli, Antonella; Moro, Stefano; Jona-Lasinio, Giovanna; Martino, Sara - 01a Articolo in rivista
rivista: FRONTIERS IN MARINE SCIENCE (ausanne : Frontiers Media S.A., 2014-) pp. - - issn: 2296-7745 - wos: WOS:000862195300001 (5) - scopus: 2-s2.0-85139126362 (8)

11573/1565763 - 2021 - Model-based clustering for estimating cetaceans site-fidelity and abundance
Caruso, Gianmarco; Panunzi, Greta; Mingione, Marco; Alaimo Di Loro, Pierfrancesco; Moro, Stefano; Bompiani, Edoardo; Lanfredi, Caterina; Pace, Daniela Silvia; Tardella, Luca; Jona Lasinio, Giovanna - 04b Atto di convegno in volume
congresso: 13th scientific meeting of the classification and data analysis group, Firenze, September 9-11, 2021 (Firenze)
libro: CLADAG 2021 book of abstracts and short papers - (978-88-5518-340-6)

11573/1644910 - 2021 - Model-based clustering for monitoring cetaceans population dynamics
Panunzi, G.; Caruso, G.; Mingione, M.; Alaimo Di Loro, P.; Moro, S.; Bompiani, E.; Lanfredi, C.; Pace, D. S.; Tardella, L.; Jona Lasinio, G. - 04f Poster
congresso: GRASPA 2021 (Rome; Italy)
libro: Graspa 2021 - (979-12-200-8496-3)

© Università degli Studi di Roma "La Sapienza" - Piazzale Aldo Moro 5, 00185 Roma