Hello, I am Louis Ohl, a happy postdoc at the Linköping University (Linköping, Sweden). My main area of interest encompasses unsupervised learning, specifically discriminative clustering, and its application with or without constraints for instance in health or material science.
I recently graduated from a Ph.D. in computer science at the Université Côte d'Azur (Nice, France) and a cotutelle Ph.D. in molecular medicine at the Université Laval (Quebec, Canada). The topic was the qualification of the aortic stenosis, a severe cardiac disease, throughout discriminative clustering methods. I had the pleasure of being supervised by Frédéric Precioso, Pierre-Alexandre Mattei in France and Arnaud Droit in Canada. If you need a quick insight on my work, you can check my vulgarisation attempt in the contest Ma thèse en 180 secondes (French only), and investigate the GemClus package that summarises most of my PhD algorithms. Prior to academia, I graduated with in computer science with a double degree: an engineering one at the Institut National des Sciences Appliquées de Lyon and at a Master's at KTH Royal Institute of Technology.
I am currently focusing on the prediction of synthesisability or dynamic stability of MXenes materials from MAX phases using ML. This type of problen is tightly linked to unsupervised methods, as it covers both PU learning (binary classification where class observation is missing) and/or semi-supervised clustering (like semi-supervised learning, but with veerrrryyy little annotated data). I work closely with Fredrik Lindsten. Do not hesitate to contact me for whatever research-related topic :-)
I am also responsible of the ML seminar series at the department of computer science (IDA) of the Linköping University.
Linköping University, Department of Computer and Information Science (IDA), Divison of Statistics and Machine Learning (STIMA)