Pierre Vandergheynst
Pierre Vandergheynst works on mathematical data processing, signals and graphs. His coauthored paper first submitted in June 2014 proposed learning graph structure so that observed signals vary smoothly across connected nodes. He became an assistant professor in January 2001 and a full professor in 2014. He became vice provost in 2015 and vice president for education in the management team starting January 2017. These historical roles connect research with university education. As of October 2026, his team studies learning methods for graphs, geometric spaces and scientific data.
Timeline
-
January 2001–2014
EPFL research appointments
Vandergheynst became an assistant professor in January 2001 and a full professor of electrical engineering in 2014. He also did postdoctoral work in signal processing.
-
June 2014
Learning graph structure
Vandergheynst coauthored work with Xiaowen Dong, Dorina Thanou and Pascal Frossard on learning graph Laplacians from signals. The model favors smooth variation across the learned graph. Its authors evaluate the approach on synthetic and real data.
-
2015
Vice-provost role
He became vice provost for education in 2015. He directed Signal Processing Laboratory 2 and researched complex, large-scale data at that time.
-
January 2017
Education leadership
He became vice president for education in a new senior-management team starting 1 January 2017. His responsibilities included teaching, doctoral education and continuing education.
-
As of October 2026
Learning on structured domains
Vandergheynst leads the LTS2 team, which develops mathematical and computational foundations for learning on graphs, geometric spaces, physical systems and scientific data. Applications include molecular biology, materials science and complex dynamical systems.
Sources and credits
Photo: Olivier Christinat; Copyright: olivier christinat. License: Not stated by source. Source: people.epfl.ch. Cropped to a square and resized. Ask us to remove or replace this photo.