Skip to content
Dr Bertrand Gauthier

Dr Bertrand Gauthier

Lecturer

School of Mathematics

Email:
gauthierb@cardiff.ac.uk
Telephone:
+44(0)29 2087 5544
Location:
M/2.31, Maths and Education Building , Senghennydd Road, Cardiff, CF24 4AG

Research Group: Statistics

Research Interests: Machine Learning, Approximation Theory, Kernel-Based Methods, Big Data, Computational Mathematics, Design of Experiments.

Previous affiliations:

  • Since January 2017: Lecturer at Cardiff University - School of Mathematics. 
  • Postdoctoral researcher at KU Leuven (Belgium), ESAT-STADIUS Centre for Dynamical Systems, Signal Processing and Data Analytics, from March 2015 to December 2016.
  • Postdoctoral researcher at CNRS - Université de Nice-Sophia Antipolis (France), Laboratoire I3S, from September 2012 to August 2014; and CNRS contract agent on secondment to EDF-Lab Chatou (France), from September to December 2014.
  • ATER (temporary research and teaching assistant) at Université de Saint-Étienne (France), Mathematics Department, Institut Camille Jordan, from February 2011 to August 2012.
  • Ph.D student and teaching assistant at École des Mines de Saint-Étienne (France) from October 2007 to January 2011.

2018

2017

2016

2014

2012

Undergraduate: 

  • Multivariate Data Analysis (Year 3, MA3506) 

Postgraduate:

  • Statistical Methods (MSc, MAT002)

My research lies at the interface between machine learning and approximation theory. Specifically, I have a genuine interest in questions related to sparsity and structures in large-scale machine problems, with a particular emphasis on sampling and super-sampling methods, model-based data-compression techniques and sparse approximation. 

Within my previous work, I investigated the use of spectral techniques for the approximation of kernel-based models and the design of experiments in second-order random-field models. I am currently studying problems related to the connections between sampling and spectral approximation of integral operators, and investigating their implications on the performance of kernel-based learning methods.