Data science & artificial intelligence
For digitalized industry & services
Each year, the University of Stanford organizes the “Women in Data Science” conference (WiDS) in 140 locations worldwide. The conference serves to inspire scientists the world over and to support women working in data science. WiDS partners Télécom Paris, Total, Inria Saclay-Île de France and Reseau InnoEnergy are organizing an online round-table discussion at this year’s conference, on 14 September 2020. The speakers at the event include Nathalie Brunelle, Project Director at Total@Saclay and President of TWICE, Assia Mouloudi, User eXperience Manager at SAP, Céline Jullien, Head of Citizen Engagement at InnoEnergy S.E and Amélie Claus, student at Télécom Paris. Florence d’Alché-Buc, who has held the DSAIDIS Chair at Télécom Paris since 2019, features in the WiDS mini-web series where women data scientists talk about their work.
In it, Florence d’Alché-Buc tells us about her background, her ambitions and the skills she has acquired, from early on in her career to the present day. She also describes the challenges posed by the machine learning of the future and its applications.
A specialist in machine learning, which is an essential data science component, she is now applying her expertise to a number of fields, including education, as she teaches IT and applied mathematics at Télécom Paris. She has set up and led a number of projects, including the Challenges program in the PASCAL network for excellence (2004-2008). Florence d’Alché-Buc has also authored around a hundred articles. She has been the scientific lead for the Digicosme Labex since 2017 and is a member of the LTCI lab at Télécom Paris.
She likes to illustrate the effectiveness of machine learning with the example of a biochemist who wants to know the exact type of a molecule in a given sample, multiple samples being available. In this scenario, machine learning will automatically find a predictive model based on sample-molecule pairs, that can be applied to other samples thanks to the computer’s ability to process regularities, thus saving the biochemist precious time.