Publications

Probabilistic Registration for Gaussian Process Three-Dimensional Shape Modelling in the Presence of Extensive Missing Data

Published in SIAM Journal on Mathematics of Data Science, 2023

We propose a method for fitting/registering shapes with missing data using Gaussian processes. Gaussian processes provide a powerful tool for shape modelling and fitting. However, existing methods in this area do not work well for detailed and deformed data with extensive missing data, such as ears. To address this, we use a multi-annotator Gaussian process regression to formulate the shape fitting problem and establish a parallel with standard probabilistic registration. Our method, called the shape fitting Gaussian process (SFGP), outperforms current approaches for registration with GP and a state-of-the-art registration method when dealing with extensive missing data. We conducted experiments on both a small 2D dataset with several transformations and a 3D dataset of ears.

Recommended citation: Valdeira, F. M., Ferreira, R., Micheletti, A., & Soares, C. (2023). Probabilistic Registration for Gaussian Process Three-Dimensional Shape Modelling in the Presence of Extensive Missing Data. SIAM Journal on Mathematics of Data Science, 5(2), 502-527.

Taxonomy for Resident Space Objects in LEO

Published in Accepted in the International Astronautical Congress 2023, 2023

Use Google Scholar for full citation

Recommended citation: Marta Guimaraes, Claudia Soares, Chiara Manfletti, "Taxonomy for Resident Space Objects in LEO." Accepted in the International Astronautical Congress 2023, 2023.

One-Shot Initial Orbit Determination in Low-Earth Orbit

Published in Accepted at the IEEE Aerospace Conference, 2023

Use Google Scholar for full citation

Recommended citation: Ricardo Ferreira, Marta Guimaraes, Filipa Valdeira, Claudia Soares, "One-Shot Initial Orbit Determination in Low-Earth Orbit." Accepted at the IEEE Aerospace Conference, 2023.

Low-Resource White-Box Semantic Segmentation of Supporting Towers on 3D Point Clouds via Signature Shape Identification

Published in arXiv preprint arXiv:2306.07809, 2023

Use Google Scholar for full citation

Recommended citation: Diogo Lavado, Cl{\'a}udia Soares, Alessandra Micheletti, Giovanni Bocchi, Alex Coronati, Manuel Silva, Patrizio Frosini, "Low-Resource White-Box Semantic Segmentation of Supporting Towers on 3D Point Clouds via Signature Shape Identification." arXiv preprint arXiv:2306.07809, 2023.

A Temporal Fusion Transformer for Long-term Explainable Prediction of Emergency Department Overcrowding

Published in Machine Learning and Principles and Practice of Knowledge Discovery in Databases: International Workshops of ECML PKDD 2022, 2022

Emergency Department overcrowding is driven, in part, by non-urgent patients that resort to emergency services despite not having a medical emergency, representing almost half of the total number of daily patients. This paper describes a novel deep learning architecture, the Temporal Fusion Transformer, that uses calendar and time-series covariates to forecast prediction intervals and point predictions for a 4-week period. We have concluded that patient volume can be forecast with a Mean Absolute Percentage Error (MAPE) of 9.87%.

Recommended citation: Caldas, Francisco M., and Cláudia Soares. "A Temporal Fusion Transformer for Long-term Explainable Prediction of Emergency Department Overcrowding." Machine Learning and Principles and Practice of Knowledge Discovery in Databases: International Workshops of ECML PKDD 2022, Grenoble, France, September 19–23, 2022, Proceedings, Part I. Cham: Springer Nature Switzerland. https://arxiv.org/pdf/2207.00610.pdf

Machine Learning in Orbit Estimation: a Survey

Published in arXiv preprint arXiv:2207.08993, 2022

This survey describes the background and motivation for the research on the Orbit prediction problem. It discusses the increasing population of human-made objects in space and the need for fast and reliable methods for space situational awareness, maneuver recommendation, and planning. The text highlights the limitations of current physics-only methods and the potential of data-driven techniques, specifically machine learning, to improve orbit prediction accuracy. It mentions the challenges related to uncertainties in the state of space objects, environmental conditions, and unknown characteristics. The text also mentions the importance of interdisciplinary research in space sciences and machine learning for physical systems. This paper provides an overview of the current work in space debris tracking and orbit prediction, emphasizing the potential of data-driven techniques, particularly machine learning, to enhance orbit prediction accuracy. It highlights the challenges associated with uncertainties and unknown characteristics of space objects and discusses the abstraction capacity of deep learning models in modeling complex non-linear systems.

Recommended citation: Francisco Caldas, Claudia Soares, "Machine Learning in Orbit Estimation: a Survey." arXiv preprint arXiv:2207.08993, 2022.

Referral prediction in Healthcare using Graph Neural Networks

Published in GReS – Workshop on Graph Neural Networks for Recommendation and Search. Co-located with the ACM RecSys ’21 conference, 2021

Access paper here

Recommended citation: REGINA DUARTE, QIWEI HAN, CLAUDIA SOARES, "Referral prediction in Healthcare using Graph Neural Networks." GReS – Workshop on Graph Neural Networks for Recommendation and Search. Co-located with the ACM RecSys ’21 conference, 2021. https://europe.naverlabs.com/wp-content/uploads/2021/09/DuarteEtAl2021.pdf