Jean-Pierre Eckmann: A geometric view of chemical reaction networks: Stability and large deviations theory in the particle approximation.

This is work with Andrea Agazzi and Amir Dembo. Chemical reaction networks
are described by polynomial vector fields, where the coordinates are the
concentrations of the chemical species. These equations are approximations
of the chemical reactions, which deal with molecules. Hence a particle
approximation in a Markov porocess. What are the probabilities
of fluctuations of this stochastic approximation? This is answered by
large deviations theory, once one establishes that it can be applied.