First year PhD Student in Grenoble.

After a Master Degree in Statistics and Data Science, I started a PhD with Sophie Achard, Julyan Arbel and Guillaume Kom Kan King in the Statify team of Inria Grenoble. My subject focus on graph inference and comparison for fMRI data using bayesian statistics. We are especially interested in quantifying uncertainty during graph analysis.

PhD Subject

Today, many data sets collect information across both space and time. A natural way to model this information is through a graph, where nodes and edges are derived from the observations. This approach is particularly useful in studying brain function, where a node represents a brain region and an edge represents a functional connection between two regions, quantified by methods such as fMRI or EEG.

To effectively use graphs for real data sets, several tasks are necessary: inference, comparison, and classification. In this proposal, we present a new unified Bayesian framework suitable for these tasks. Graphs are constructed by estimating a multivariate density with a density per edge. For comparisons, distances between these graphs are developed to account for the uncertainty quantified during inference. Finally, classification methods are designed for various applications, including the clinical explanation of brain disorders.

Interests

I am interested in application in neurosciences, biology and sociology. I focus on interpretability and robustness of models.