Guglielmo Ferranti

Guglielmo Ferranti is a theoretical physicist currently enrolled as a Ph.D. student in Complex Systems at the Università degli Studi di Catania. His academic journey began with a focus on Physics that was later shifted on exploiting the growing potential of Machine Learning and Data Science techniques to solve challenging problems. Guglielmo's expertise spans theoretical and applied physics, mathematics, statistical analysis, data science, and machine learning. He has developed and tested various machine learning models, demonstrating proficiency in extracting valuable information from complex datasets.

His Ph.D. work includes:

1. Earthquake Risk Assessment (Università degli Studi di Catania): Utilized a combination of Artificial Neural Networks and Random Forest approaches to analyze structural vulnerability in buildings affected by the 2009 L’Aquila earthquakes. Employed explainable AI techniques to derive a continuous vulnerability metric, outperforming traditional risk assessment methods.

2. Texts-as-Graphs Classification (Università degli Studi di Catania): Developed a method to represent texts as networks for classifying sentences from different authors based on the topology of the respective graph, using Graph Convolutional Networks (GCNs) and Node2vec.

3. Road Infrastructure Vulnerability (ANAS or Azienda Nazionale Autonoma Strade Statali): Partnered with ANAS to evaluate road sections by integrating vulnerability and topological significance through network science techniques.

During his Master’s program, he collaborated with the Nuclear Physics Laboratory (INFN) to develop a machine learning approach for simulating the transport of a beam of particles through a series of magnetic lenses, significantly reducing simulation time.

Additionally, Guglielmo has been involved in medical data science, collaborating with neurosurgeons at San Marco Hospital to forecast brain movement during cranial surgery (Brain-Shift problem) and applying UMAP techniques to cluster patients for cancer survivability studies.

Guglielmo's multidisciplinary work reflects a unique blend of physics, machine learning, and data science, demonstrating his ability to apply advanced mathematical and computational techniques to a diverse range of real-world problems.