Welcome to Sylvain Combettes' homepage!

  I am a first-year PhD student in Data Science at the Borelli Center, a research department of École Normale Supérieure Paris-Saclay. I am advised by Laurent Oudre and Charles Truong.
I am focusing on a symbolic representation for physiological signals (biomedical time series). Research interests: Representation learning, Symbolic approaches, Change-point detection, Pattern recognition.

Details.The objective of the thesis is to construct symbolic representations adapted for the processing of physiological data acquired during multi-sensor experimental protocols. The analysis of these heterogeneous data is particularly complex because of their multivariate, multimodal and non-stationary properties. The idea is to represent a data flow in the form of a series of symbols (barcodes), reflecting not only the temporal evolution of the observed phenomena, but also the evolution of the link between the different modalities. These representations will be based on the study of local phenomena observed in the data (stationary regimes or repetitive patterns) in order to adapt the representation to the physiological phenomena of interest. One of the stakes of this thesis is also to build adapted metrics allowing to compare these data, especially for longitudinal monitoring and inter-individual comparison.

  I share some of my Data Science projects on my GitHub: codes, reports, synthetic reports and slides.

  On my blog, I mainly post about artificial intelligence and machine learning on both technical and non-technical aspects.

  sylvain.combettes [a t] ens-paris-saclay.fr

AI

Photo by Franki Chamaki from Unsplash.