Gautier Appert
Research Scientist
GitHub - LinkedIn - Lichess
gautier [dot] appert [dot] chess [at] gmail [dot] com
I am a French PhD Candidate in Statistics at CREST-ENSAE, under the direction of Dr. Olivier Catoni (CNRS Research Director). I am mainly working on various type of k-means and PAC-Bayesian theory with application to images.
I graduated from ENSAE ParisTech, completing an Engineer’s degree in Data Science, and from ENS Paris-Saclay, completing a MSc in Mathematics, Vision and Machine Learning. I also hold a MSc in Statistics and Econometrics from Toulouse School of Economics.
You can find my PhD dissertation Information k-means, fragmentation and syntax analysis: A new approach to unsupervised Machine Learning and my PhD defense presentation (Slides)
Full R and C++ source code for the fragmentation and syntax analysis algorithm is available on GitHub
Presentation of a new type of k-means using the Kullback divergence as a distortion measure, with application to digital images (Slides)
2020-2021, as a Temporary Teaching and Research Assistant in Statistics at University of Paris 1 Pantheon-Sorbonne
2019-2020, as a Teaching Assistant (“ATER” position) at Paris-Saclay University
2016-2019, as a Teaching Assistant at ENSAE Paris Tech (list of courses)