Most neural network learning algorithms cannot use knowledge other than what is provided in the training data. Initialized using random weights, they cannot use prior knowledge such as knowledge stored in previously trained networks. This manuscript thesis addresses this problem.…

We present a novel approach to simulate the motion of vortex filaments in three dimensional, incompressible flows. We represent vortex filaments as thin tubes on a 3D-grid and use a Lagrangian vortex filament method to describe their motion within a…