Abstract

We consider initial value problems of nonlinear dynamical systems, whichinclude physical parameters. A quantity of interest depending on the solutionis observed. A discretisation yields the trajectories of the quantity ofinterest in many time points. We examine the mapping from the set of parametersto the discrete values of the trajectories. An evaluation of this mappingrequires to solve an initial value problem. Alternatively, we determine anapproximation, where the evaluation requires low computation work, using aconcept of machine learning. We employ feedforward neural networks, which arefitted to data from samples of the trajectories. Results of numericalcomputations are presented for a test example modelling an electric circuit.

 

Quick Read (beta)