de Hemptine b.234
010 47 3265
View my 'Don't tell people how to do things. Tell them what to do and let them surprise you with their results. '
-- G. Patton
: Who I am . .
I am sysadmin at the Supercomputing centre of the Université catholique de Louvain. I have got a M.Sc. in Computer Science and Engineering and a Ph.D. in Applied Mathematics. I was for 8 years a researcher in machine learning a the Department of Mathematical Engineering, and a member of the CESAME Research Center and of the Machine Learning Group at UCL.
: What I do . .
As a member of the sysadmin team, I set up and maintain computer clusters dedicated to research and scientific computing. I also provide user support for developing/porting/maintaining their source codes.
My field of research was machine learning. Machine learning is a subfield of artificial intelligence concerned with how computer programs can evolve and get better and better as they gain experience ('learn'). The typical software example is the spam filter in your favorite email client. At the beginning, it makes a few mistakes, and you have to correct it. But as time goes, you need less and less to correct it because it has learned in a sense your particular definition of spam.
I more particularly focussed on statistical learning, with emphasis on data mining applications. The aim of data mining is to 'find patterns', or 'gain knowledge' from datasets. These patterns often appear as relationships between data (clustering: homogeneous groups of similar data), or between data attributes (classification/regression: prediction of one of the attributes -- the target-- based on the others).
The tools I specialized in are similarity-based models and algorithms, and kernel methods. My research interests lie in feature selection and meta learning. I have mainly played with spectral data, face images, genomic data, and econometric data.