- 400 Downloads
In the introduction to this thesis, some time was spent discussing the choices one makes in modelling any system. In some sense we are playing a game by which we wish to incorporate enough detail so as to be realistic and informative, but not so much as to render the model resistant to interpretation. Having struck the balance between these competing considerations, we may be further confounded by the analytic intractability of the resulting problem. While the stochastic nature of the Moran model, makes it difficult to solve in its entirety, its one-dimensional nature makes other quantities, such as the fixation probability and fixation time, obtainable. However many models inspired by nature (especially those which are nonlinear and in many dimensions) stubbornly resist analytic treatment.
KeywordsFull System Fast Variable Noise Covariance Matrix Moran Model Conditioning Method