I have been interested in various research topics for some time now, first as a hobby, then since 1999 as my main occupation at EDF R&D. Beside research, I also teach statistics and the use of SAS.
I'm currently doing a PhD thesis with Pr. Minoux (University Paris 6) on stochastic mixed-integer linear programming applied to nuclear plant maintenance scheduling.
I'm also responsible for the Stochastic Optimization project at EDF R&D.
Two other colleagues (K. Barty and C. Strugarek) and I, proposed in 2004 a stochastic gradient algorithm in Hilbert spaces that is implementable. We tried it with success on a few small closed loop problems and extended our approach to stochastic dynamic programming problems. We then applied it to option pricing with the help of P. Girardeau, during his intership at EDF R&D.
The following talks were given by my colleagues on our common research subjects.
This is the main subject of my ongoing thesis with Pr. Minoux. I'm currently mainly focused on disjunctive programming and problem reformulation.
As my education was mainly centered on economy, statistics and computing, my first research subject was naturally applied statistics. After some small econometric work for the french welfare system, I mainly focused on electricity markets and more precisely on price and load forecasting. I did one of the first spot price model at EDF with three colleagues (H. Fedry, T. Socroun and F. Dazy). After that, I redesigned the french load forecasting model twice: the first model is operationnal at EDF Trading since a few years, the second one is more an advanced proof of concept. Sadly, most of this work was quite confidential.
The following talks were given by my colleagues on our common research subjects.