Home

ISABELLE GUYON

Isabelle Guyon

Director, Research Scientist, Google DeepMind (since Oct 2022).

Chaired Professor of Artificial Intelligence (PR EX1) and INRIA researcher, Machine Learning and Optimization (TAU team), Laboratoire Interdisciplinaire des Sciences du Numérique (LISN) University Paris-Saclay, France

Graduate of ESPCI. PhD in Paris in Gerard Dreyfus' Lab. Worked at Bell Labs for 7 years, then moved to Berkeley, California. Three children with Bernhard Boser. Independent consultant with ClopiNet

Since 2015, Full Professor at Université Paris-Saclay:  Teaching Machine Learning and advising graduate students. 2019-2021, coordinator of CS Artificial Intelligence master program at UPSaclay.

Co-program chair of NeurIPS 2016 and co-general chair of NeurIPS 2017. [what the press says]; then NeurIPS board member. AMIA and an ELLIS fellow. Action editor at JMLR, and CiML springer series editor. BBVA award recipient (2020).

Recent awards: Ten-Year Technical Impact Award ICMI, 2023; Best paper award 19th IEEE conference on e-Science, 2023.

President of ChaLearn

Bear Valley, California

501(c)(3) non-profit organizing challenges

Since 2003 machine learning challenge organizer. Using challenges as a means of directing research in domains including: causality, computer vision, automatic machine learning, and high energy physics.

ChaLearn collaborates with the open-source Codalab competitions project , of which I am community lead.

We received a large EU grant to organize the 2 million Euro See.4C challenge with data from RTE. The winners have been announced. We also co-organized two AI for Industry challenges with Dassault Aviation and RTE France, funded by Région Ile-de-France. See the results of the latest L2RPN challenge.

Our current projects include Looking at People,  TrackML, Meta-Learning challenges, and Learning to run a Power Network, and systematic bias in computer vision, medical data, and High Energy Physics (HEP).

March 2024: ML contests announces that Codalab is the most used ML competition platform. Codalab competitions is an open-source project of which Isabelle Guyon is community lead  [Image credit: Harald Carlens].

March 2023:  Exposition Femmes de Sciences, 140 ans de l'ESPCI. [Image credit: Anthony Thornton]

December 2022: NeurIPS keynote on datasets and benchmarks: How ML is becoming an experimental science. [Image credit: Dall-E]

August 2022: We received with Lawrence Berkeley Labs a grant of 6.4 million USD to develop benchmarks in High Energy Physics and implement them on Codabench. [Image credits: X-ray by NASA/CXC/PSU/L.Townsley et al; Optical by UKIRT; Infrared by NASA/JPL-Caltech]

We are starting a new ANR Artificial Intelligence Chaire

April 2020: AutoDL challenge design and results (NAS workshop @ ICLR 2020) [SLIDES] ( Talk in Google drive good resolution OR [Youtube part I] [Youtube part II] ). A SERIES sponsored by ChaLearn, Google, 4Paradigm. We participated to the Jean Zay grand challenge (1000 GPU super-computer).

SVM

The invention of SVMs

 (the story KDNuggets 2016 and eye-on.ai podcast, 2023 )

https://www.dropbox.com/s/t6tns9yfrnc1u2p/Democratiser_IA.pdf?dl=0

Avril 2018 - Article dans Le Monde:  Isabelle Guyon veut democratiser l'intelligence artificielle. [Image credit: Marion Berrin]

National Academy of Sciences

March 2019: Sackler Coloquium of the Science of Deep Learning. A presentation of the work done in collaboration with RTE France to help predict the flow of electricity in the French power grid.


Academie des Sciences

November 2018: La Science Hors Murs. I was invited to talk about my journey as challenge organizer and how to organize contests of contests.