Filippo Guerranti

Hello everybody, I’m Filippo! Welcome to my personal page 👋

I’m a PhD student in Data Analysis and Machine Learning at the Technical University of Munich under the supervision of Prof. Stephan Günnemann. My interests lie in the fields of machine learning and deep learning for graph-structured data.

I received my master’s degree in Computer and Automation Engineering at the Department of Information Engineering and Mathematics with a thesis on relational reasoning learning. The thesis is titled Approximating Logical Reasoning using Neural Network and aims to enable machines to reason about related entities using a priori knowledge. I was also a member of the Siena Artificial Intelligence Lab (SAILab), University of Siena, lead by Prof. Marco Gori.

My goal as a PhD student is to study and deepen my knowledge to become an independent researcher and to create machine learning models that are reliable, robust under uncertainties, and fair. This will enable us to safely use them in sensitive and critical real-world situations without losing efficiency and performance.

In my spare time I enjoy CrossFit, running, hiking, mountain biking in the woods and hills of Tuscany (now in the woods and hills of Munich) and, occasionally, bouldering (I’m still a beginner). At certain times of the year I’m also an avid reader (books on psychology, sociology, popular science and novels), at other times I just can’t open a book.

Research interests: graph structure learning, graph neural networks, graph representation learning, relational reasoning, interpretability & reliability of ML models.

News

Jun 13, 2022	I started my PhD in the Data Analysis and Machine Learning group at Technical University of Munich. I look forward to this new life and professional adventure!
Apr 26, 2022	I graduated as a MSc in Computer and Automation Engineering at the University of Siena with the final grade of 110/110 cum laude with Honours (highest possible grade). The thesis is entitled Approximating Logical Reasoning using Neural Network.
Mar 22, 2022	I have been accepted as a PhD student in Data Analysis and Machine Learning group at Technical University of Munich. I will be joining Prof. Stephan Günnemann group at the beginning of June 2022.
Jan 24, 2022	I presented my Master’s thesis to the Data Analysis and Machine Learning group at the Technical University of Munich.
Jan 19, 2022	I created this site! Let’s see what comes next 🤯.

Selected publications

KES

CaregiverMatcher: graph neural networks for connecting caregivers of rare disease patients

Guerranti Filippo, Mannino Mirko, Baccini Federica, Bongini Pietro, Pancino Niccoló, Visibelli Anna, and Marziali Sara

Procedia Computer Science 2021

Abs Bib HTML PDF

Rare diseases affect a growing number of individuals. One key problem for patients and their caregivers is the difficulty in reaching experts and associations competent on a particular disease. As a consequence, caregivers, often family members of the patient, learn much about the disease from their own experience. CaregiverMatcher is a proof of concept providing a smart solution to build a network of caregivers, linked by a matching mechanism based on graph neural networks. The caregivers and their experience with rare diseases are described by node features. Associations and care centers are invited to share their knowledge on the platform.

@article{guerranti2021caregivermatcher,
                 title     = {CaregiverMatcher: graph neural networks for connecting caregivers of rare disease patients},
                 journal   = {Procedia Computer Science},
                 volume    = {192},
                 pages     = {1696-1704},
                 year      = {2021},
                 note      = {Knowledge-Based and Intelligent Information & Engineering Systems: Proceedings of the 25th International Conference KES2021},
                 issn      = {1877-0509},
                 doi       = {https://doi.org/10.1016/j.procs.2021.08.174},
                 url       = {https://www.sciencedirect.com/science/article/pii/S1877050921016690},
                 author    = {Guerranti, Filippo and Mannino, Mirko and Baccini, Federica and Bongini, Pietro and Pancino, Niccoló and Visibelli, Anna and Marziali, Sara},
                 keywords  = {Cross-platform application, Caregivers, Rare diseases, Graph Neural Networks, Deep Learning},
                 abstract  = {Rare diseases affect a growing number of individuals. One key problem for patients and their caregivers is the difficulty in reaching experts and associations competent on a particular disease. As a consequence, caregivers, often family members of the patient, learn much about the disease from their own experience. CaregiverMatcher is a proof of concept providing a smart solution to build a network of caregivers, linked by a matching mechanism based on graph neural networks. The caregivers and their experience with rare diseases are described by node features. Associations and care centers are invited to share their knowledge on the platform.}
                 }