I’m a french Ph.D. student in machine learning applied to medical imaging. My favorite languages are Python 3 and mathematics and I believe in open source and public data. Feel free to contact me at firstname.lastname@example.org.
My ORCID: 0000-0002-5725-7910
Nov. 2018–Now! Cifre Ph.D. student. QuantIF and Henri Becquerel Center and AQUILAB. France. Topic—Prediction of survival and adverse effects of (chemo)radiotherapy in lung and head-and-neck cancers. Application of machine learning to medical imaging.
March–Aug. 2018. Research Intern. INRIA Lille Nord Europe, MAGNET team. Lille, France. Topic—Improvement of GOPA, a privacy-preserving algorithm for distributed averaging over P2P networks. Robustness was provided against users disconnections and malicious adversaries.
Jan. 2017–Jan. 2018. Lead Python Developer @ ISO9001:2015 Certified INSA Project, Unicancer Innovation Award 2017: Management and Research Award. Henri Becquerel Center and INSA Rouen Normandie, Dept. Information Systems Architectures. Rouen, France. Topic—Conception and development of a Python GUI for the segmentation of cancer patients’ CT scans. Automation of the process thanks to Keras-implemented deep and fully convolutional networks.
June–Aug. 2017. R&D Data Scientist Intern. Vekia. Lille, France. Topic—Development of a Python module for detecting suspicious stocks—phantom inventories—with machine learning techniques.
July 2013. Production Workman Intern. Ressorts Masselin. Le Petit-Quevilly, France. Topic—Cold springs rolling, grinding, shot peening, quality controlling and packaging.
July 2012. French Baccalaureate of Sciences, Magna Cum Laude, Major in Earth and Life Sciences. French School Abroad Montaigne. Cotonou, Benin.
Oct. 2017.Unicancer Innovation Award 2017: Management and Research Award. Unicancer. France.
May 2017.4th and Best Presentation Award @ Speed Data Scientist 2017: Failures detection challenge. Société Générale. Paris, France.
Nov. 2017– gflsegpy: A Python 3 implementation of the group fused Lasso for multiple change-point detection. This project implements the group fused Lasso (GFL), as defined by Bleakley and Vert, 2011, a method for detecting breakpoints in multidimensional signals. The gflsegpy package provides an implementation of two GFL algorithms plus a demo script and a data visualization module.
Jan.–June 2016. For students or teachers: Scientific and Technological Support in Primary School: An introduction to robotic programming for children.