Marco Canducci
Office 206, School of Computer Science, University of Birmingham, Edgbaston, B15 1PQ, UK
m.canducci@bham.ac.uk •
@CanducciMarco •
LinkedIn •
GitHub
Marie Skłodowska-Curie Actions Individual Postdoctoral Fellowship, European Commission (2025)
Project HARMONISE received a score of 90/100 in the MSCA-PF 2025 call, exceeding the evaluation threshold and qualifying for the Seal of Excellence. The project proposes the probabilistic homogenisation of stellar photosphere parameter spaces estimated from multiple astronomical surveys, through shared latent stochastic process representations. Not funded due to budget constraints.
Research Interests
Probabilistic machine learning, manifold learning, interpretable computer vision, swarm-intelligence-based algorithms, discriminative subspace methods. Applications in computational astrophysics (galaxy structure, stellar streams, cosmic filaments), medical informatics (multi-morbidity, disease progression), and materials science (surrogate modelling for aerospace manufacturing).
Employment
School of Computer Science, University of Birmingham, UK
Development of surrogate models for directional solidification of metallic superalloys for aerospace turbine components. Research in Computer Vision, Machine Learning, and AI (unsupervised, supervised, self-supervised learning).
School of Computer Science, University of Birmingham, UK
Development of digital tools for multi-morbidity investigation in patients. Epidemiological sequence interpretation across diverse demographics. Visual representation methods for polypharmacy risk in complex multi-morbidity patients.
School of Computer Science, University of Birmingham, UK
Feature importance estimation in high-dimensional datasets for Cushing’s disease progression from metabolic data. Classification and dimensionality reduction with discriminant metric tensor. Probabilistic GMLVQ formulation for datasets with measurement error.
Education
University of Birmingham, UK
Part of the MSCA-ITN-ETN SUNDIAL network. Research on probabilistic manifold learning and its applications to astrophysical datasets.
Università Alma Mater Studiorum di Bologna, Italy
Università Alma Mater Studiorum di Bologna, Italy
Selected Publications
For the full list see the Publications page.
A. Prete, M. Canducci et al. Endocrine and metabolic determinants of cardiometabolic risk in mild autonomous cortisol secretion. eBioMedicine (The Lancet), 2026.
R. Baier-Soto et al. incl. M. Canducci. The role of supercluster filaments in shaping galaxy clusters. A&A 704 A228, 2025.
L. Spina et al. incl. M. Canducci. Deep chemical tagging with graph attention networks. A&A 702 A267, 2025.
P. Awad, M. Canducci et al. Swarming in stellar streams: Unveiling the Jhelum stream with ant colony-inspired computation. A&A 683, 2024.
M. Canducci et al. 1-DREAM: 1D Recovery, Extraction and Analysis of Manifolds in noisy environments. Astronomy & Computing 41, 2022.
M. Canducci, P. Tiňo, M. Mastropietro. Probabilistic modelling of general noisy multi-manifold data sets. Artificial Intelligence 302, 2022.
Teaching
Student Supervision
PhD: Othman Alghamdi (2023–), Meerah Al-Hakbani (2023–), Dario Barone (2024–)
Master’s: Steni Sebastian (2020) — Gravitational wave population parameters via Probabilistic Hough Transform
Bachelor’s: Christopher Callum Nailer (2022), Jiahao Gai (2022)
Selected Conferences & Invited Talks
Academic Service
Peer reviewer: IEEE Transactions on Pattern Analysis and Machine Intelligence • Neural Networks • Pattern Recognition • Astronomy & Astrophysics • AAAI • ACM KDD
Skills
Programming: Python, MATLAB, C
Languages: Italian (native), English (C2)