Giuseppe Prencipe

Prof. Giuseppe Prencipe

UNIVERSITA' DI PISA, Pisa, Tuscany, Italy
orcid

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
29 November 2016 in Distributed Computing
Verified
Journal Article
Closed access
PDF file
PFPaola Flocchini
Paola Flocchini
Giuseppe PrencipeGiuseppe Prencipe
Giuseppe Prencipe
NSNicola Santoro
Nicola Santoro
Giovanni VigliettaGiovanni Viglietta
Giovanni Viglietta

26 March 2024 in Artificial Intelligence in Medicine
Verified
Journal Article
Open access
TTTommaso Turchi
Tommaso Turchi
Giuseppe PrencipeGiuseppe Prencipe
Giuseppe Prencipe
AMAlessio Malizia
Alessio Malizia
SFSilvia Filogna
Silvia Filogna
FLFrancesco Latrofa
Francesco Latrofa
GSGiuseppina Sgandurra
Giuseppina Sgandurra
The ongoing digital revolution in the healthcare sector, emphasized by bodies like the US Food and Drug Administration (FDA), is paving the way for a shift towards person-centric healthcare models. These models consider individual needs, turning patients from passive recipients to active participants. A key factor in this shift is Artificial Intelligence (AI), which has the capacity to revolutionize healthcare delivery due to its ability to personalize it. With the rise of software in healthcare and the proliferation of the Internet of Things (IoT), a surge of digital data is being produced. This data, alongside improvements in AI's explainability, is facilitating the spread of person-centric healthcare models, aiming at improving health management and patient experience. This paper outlines a human-centered methodology for the development of an AI-as-a-service platform with the goal of broadening access to personalized healthcare. This approach places humans at its core, aiming to augment, not replace, human capabilities and integrate in current processes. The primary research question guiding this study is: "How can Human-Centered AI principles be considered when designing an AI-as-a-service platform that democratizes access to personalized healthcare?" This informed both our research direction and investigation. Our approach involves a design fiction methodology, engaging clinicians from different domains to gather their perspectives on how AI can meet their needs by envisioning potential future scenarios and addressing possible ethical and social challenges. Additionally, we incorporate Meta-Design principles, investigating opportunities for users to modify the AI system based on their experiences. This promotes a platform that evolves with the user and considers many different perspectives.

29 June 2022 in Sensors
Verified
Journal Article
Open access
PDF file
DADaniele Atzeni
Daniele Atzeni
DBDavide Bacciu
Davide Bacciu
DMDaniele Mazzei
Daniele Mazzei
Giuseppe PrencipeGiuseppe Prencipe
Giuseppe Prencipe
Wireless networks have drastically influenced our lifestyle, changing our workplaces and society. Among the variety of wireless technology, Wi-Fi surely plays a leading role, especially in local area networks. The spread of mobiles and tablets, and more recently, the advent of Internet of Things, have resulted in a multitude of Wi-Fi-enabled devices continuously sending data to the Internet and between each other. At the same time, Machine Learning has proven to be one of the most effective and versatile tools for the analysis of fast streaming data. This systematic review aims at studying the interaction between these technologies and how it has developed throughout their lifetimes. We used Scopus, Web of Science, and IEEE Xplore databases to retrieve paper abstracts and leveraged a topic modeling technique, namely, BERTopic, to analyze the resulting document corpus. After these steps, we inspected the obtained clusters and computed statistics to characterize and interpret the topics they refer to. Our results include both the applications of Wi-Fi sensing and the variety of Machine Learning algorithms used to tackle them. We also report how the Wi-Fi advances have affected sensing applications and the choice of the most suitable Machine Learning models.

16 November 2020 in IEEE Access
Verified
Journal Article
Open access
Gianlorenzo D'AngeloGianlorenzo D'Angelo
Gianlorenzo D'Angelo
MDMattia D'Emidio
Mattia D'Emidio
SDShantanu Das
Shantanu Das
ANAlfredo Navarra
Alfredo Navarra
Giuseppe PrencipeGiuseppe Prencipe
Giuseppe Prencipe
We study models and algorithms for Programmable Matter (PM), that is matter with the ability to change its physical properties (e.g., shape or optical properties) in a programmable fashion. PM can be implemented by assembling a system of weak self-organizing computational elements, called particles , that can be programmed via distributed algorithms to collectively achieve some global task. Recent advances in the production of nanotechnologies have rendered such systems increasingly possible in practice, thus triggering research interests from many areas of computer science. The most established models for PM assume that particles: are modeled as finite state automata; are all identical, executing the same algorithm based on local observation of the surroundings; live and operate in the cells of a hexagonal grid; can move from one cell to another by repeatedly alternating between a contracted state (a particle occupies one cell) and an expanded state (a particle occupies two neighboring cells). Given these elementary features, it is rather hard to design distributed algorithms even for basic tasks and, in fact, all existing solutions to solve fundamental problems via PM have resorted to endowing PM systems with various capabilities to overcome such hardness, thus assuming quite unrealistic features. In this paper, we move toward more realistic computational models for PM. Specifically, we first introduce ${\sf SILBOT}$ , a new modeling approach that relaxes several assumptions used in previous ones. Second, we present a distributed algorithm to solve, in the ${\sf SILBOT}$ model, a foundational primitive for PM, namely Leader Election . This algorithm works in $O(n)$ rounds for all initial configurations of $n$ particles that are both connected (i.e. particles induce a connected graph) and compact (i.e. without holes , that is no empty cells surrounded by particles occur). As usual in asynchronous contexts, a round is intended as the time within which all particles have been activated at least once. Third, we show that, if the initial configuration admits holes, it is impossible to achieve leader election while preserving connectivity. Finally, by slightly empowering the robots, we design an algorithm to handle initial configurations admitting holes that in $O(n^{2})$ rounds solves the leader election problem while obtaining also compaction.

23 March 2024 in International Journal of Cardiology
Verified
Journal Article
Open access
APAlessandro Pingitore
Alessandro Pingitore
CZChenxiang Zhang
Chenxiang Zhang
CVCristina Vassalle
Cristina Vassalle
PFPaolo Ferragina
Paolo Ferragina
PLPatrizia Landi
Patrizia Landi
ASAlina Sîrbu
Alina Sîrbu
Machine learning (ML) employs algorithms that learn from data, building models with the potential to predict events by aggregating a large number of variables and assessing their complex interactions. The aim of this study is to assess ML potential in identifying patients with ischemic heart disease (IHD) at high risk of cardiac death (CD). 3987 (mean age 68 ± 11) hospitalized IHD patients were enrolled. We implemented and compared various ML models and their combination into ensembles. Model output constitutes a new ML indicator to be employed for stratification. Primary variable importance was assessed with ablation tests. An ensemble classifier combining three ML models achieved the best performance to predict CD (AUROC of 0.830, F1-macro of 0.726). ML indicator use through Cox survival analysis outperformed the 18 variables individually, producing a better stratification compared to standard multivariate analysis (improvement of ∼20%). Patients in the low risk group defined through ML indicator had a significantly higher survival (88.8% versus 29.1%). The main variables identified were Dyslipidemia, LVEF, Previous CABG, Diabetes, Previous Myocardial Infarction, Smoke, Documented resting or exertional ischemia, with an AUROC of 0.791 and an F1-score of 0.674, lower than that of 18 variables. Both code and clinical data are freely available with this article. ML may allow a faster, low-cost and reliable evaluation of IHD patient prognosis by inclusion of more predictors and identification of those more significant, improving outcome prediction towards the development of precision medicine in this clinical field.

12 May 2020 in IEEE Access
Verified
Journal Article
Open access
SDShantanu Das
Shantanu Das
PFPaola Flocchini
Paola Flocchini
Giuseppe PrencipeGiuseppe Prencipe
Giuseppe Prencipe
NSNicola Santoro
Nicola Santoro