“I believe that projects like this one related to the use of data in the healthcare field can have a high impact. The predictions obtained through mathematical models of machine learning can be support for decision making in medical prognosis or diagnosis

• PhD student Jose Ignacio Segovia Martín will participate in the ICML congress together with his supervisor Dr. Santiago Mazuelas with the paper "Double weighting for covariate shift adaptation"

The scientific article "Double weighting for covariate change adaptation" is part of the project "Early prognosis of COVID-19 infections using machine learning" in which the Basque Center for Applied Mathematics - BCAM participates and it will end in September 2023. This project is funded by AXA Research Fund and is part of the exceptional call for proposals Risk mitigation after the COVID-19 pandemic.

The project is in its final phase and after several articles related to the project, the last one has been published by Machine Learning PhD student Jose Ignacio Segovia Martín and his supervisor Dr. Santiago Mazuelas. "In Double weighting for covariate change adaptation, we deal with the COVID-19 problem from the situation where the patient data collected during the learning stage and the patient data we want to predict belong to different geographical regions or waves", says Segovia.

Segovia recently presented a poster at the 5th edition of the BIDAS Congress at BCAM. They will also present the paper at the International Conference on Machine Learning (ICML) and at the SEIO national congress.

This work has led to the collaboration with Professor Angi Lui from Johns Hopkins University and Jose Ignacio, will spend 3 months in the Computer Science department with the researcher.

The project "Early prognosis of COVID-19 infections using machine learning" aims to develop machine learning techniques that predict the future severity of infections using health data obtained at the time of infection detection. " For instance, an infected patient with a negative (resp. positive) early prognosis can be directly transferred to semi-intensive care (resp. regular ward) before he/she undergoes notable symptoms. In addition, the prediction algorithms developed in the project can also be used to closely monitor not infected individuals with high probabilities of being asymptomatic or suffer complications in case they become infected by COVID-19," says the PhD student.

"I believe that projects like this one related to the use of data in the health field can have a high impact on society. The predictions obtained through mathematical models of machine learning can support decision-making in prognosis or medical diagnoses," concludes Segovia.

About Jose Ignacio Segovia Martín

Graduated in Mathematics from the University of Valladolid (2015-2019) and finished his degree with the Final Project "Introduction to the numerical integration of algebraic differential equations". During his studies, he moved thanks to an Erasmus + scholarship to the University of Opole (Poland) in the academic year 2018-2019. He studied the master’s degree in mathematics research at the University of Valladolid and her master’s Thesis was entitled "Network Coding and error coding" and obtained an Honours degree. During his master’s studies he obtained a grant from the Social Council to collaborate in research tasks in departments and L.O.U. institutes of the University of Valladolid. Segovia, is currently a PhD student in Computer Engineering at the UPV/EHU at the Basque Center for Applied Mathematics - BCAM, under the direction of Dr. Santiago Mazuelas with the title "Minimax supervised classification with applications to COVID-19 prognosis".