Paula Gordaliza Pastor

Postdoctoral Fellow

Photo not available

T +34 946 567 842
F +34 946 567 843

Information of interest

Artificial Intelligence technologies are undoubtedly making human life easier over the last years. In particular, machine learning based systems are reaching society at large and in many aspects of the everyday life and the professional world. Yet with its benefits, machine learning techniques are not absolutely objective since model classifications and predictions rely heavily on potentially biased data. Hence this generalization of predictive algorithms has been accompanied by concerns about the ethical issues that may arise from the adoption of these technologies, not only among the research community but also among the entire population. Thanks to this, there has been a great push for the emergence of multidisciplinary approaches for assessing and removing the presence of bias in machine learning algorithms.
The purpose of my doctoral thesis, entitled "Fair Learning: an optimal transport based ap- proach", was presenting a mathematical approach for the fairness problem in machine learning. The application of our theoretical results aimed at shedding some light on the maelstrom of techniques or mere heuristics that ceaselessly appear to address these issues. We believe that a robust mathematical ground is crucial in order to guarantee a fair treatment for every subgroup of population, which will contribute to reduce the growing distrust of machine learning systems in the society.

I completed this doctoral thesis project last September 2020 and began right after a postdoctoral fellowship at BCAM where I am involved in the project named "Fair Learning in Health", that so much could describe a continuation of my PhD research.

I am also working at the University of Valladolid as an Associated Professor since November 2020.

  • A survey of bias in machine learning through the prism of statistical parity 

    Besse, P.; del Barrio, E.; Gordaliza, P.Autoridad BCAM; Loubes, J.-M.; Risser, L. (2020)
    Applications based on machine learning models have now become an indispensable part of the everyday life and the professional world. As a consequence, a critical question has recently arose among the population: Do algorithmic ...

More information