### Data Scientist at Qwant (Nice, France) (2019 - now)

I am a data scientist engineer at Qwant . As a member of the team "Search", our goal is to improve the quality of the results delivered by www.qwant.com using Learning To Rank algorithms.

I am a data scientist engineer at Qwant . As a member of the team "Search", our goal is to improve the quality of the results delivered by www.qwant.com using Learning To Rank algorithms.

I did my postdoc at Universitat Pompeu Fabra in Barcelona (Spain) in the AI-ML team in Reinforcement Learning. I was involved in the European Project * Dynamically Evolving Long-Term Autonomy * (DELTA).

We applied (Hierarchical) Reinforcement Learning to optimize the electrical consumption of a smartgrid. We worked on a microgrid simulator developed by Montefiore research unit in Liège (Belgium).

I developed this webpage for dentists.

I finished my PhD in 2018 at Aix-Marseille Université in Mathematics in Control Theory of Partial Differential Equations. I studied control properties of continuous and discretized Systems of Parabolic Differential Equations (PDE) (for instance, the heat equation \(\partial_t f -\Delta f = 0\) is a Parabolic PDE) and continuous Degenerated Parabolic Equations (More precisely, Grushin's Equation \(\partial_t f - \partial_{xx} f - |x|^{2 \gamma} \partial_{yy} f= 0 \)) . The spectral properties of Sturm-Liouville Differential Equations were crucial in the analysis. Find my thesis here.

See the introduction of my PhD for a detailed presentation.

We developed the following web application using HTML, CSS, Tapestry, Javascript and PHP.

I also applied our ideas on Atari Games environments and more specifically on the hard exploration game Montezuma's Revenge.

Maxime Euziere and I participated to the 2kPlus Challenge.
We finished 6th out of 20 participants with a demake of the famous flash game SHIFT (if you don't know Shift, look at this video)

Try out our shift demake written is JS in only 2ko ! (GitHub repo here)

Thanks to Daniel Shiffman for his code available on github (coding challenge 115).

Click to play Snake ! (GitHub repo here)

Click here to draw something ! (GitHub repo here )