Kurzfassung

SEEDS: Structural Error Estimation in Dynamic Systems Models

Dynamical models are useful tools for understanding, predicting and controlling complex bio- logical systems. Paradoxically, the biological complexity and the incompleteness of our knowl- edge is also the main obstacle for model development. In this interdisciplinary project we will devise computational methods to deal with incomplete and uncertain information in dynamic biological models. We will investigate fundamental limitations...SEEDS: Structural Error Estimation in Dynamic Systems Models

Dynamical models are useful tools for understanding, predicting and controlling complex bio- logical systems. Paradoxically, the biological complexity and the incompleteness of our knowl- edge is also the main obstacle for model development. In this interdisciplinary project we will devise computational methods to deal with incomplete and uncertain information in dynamic biological models. We will investigate fundamental limitations to the observability of systematic or structural model errors and design more accurate and efficient algorithms for the simultaneous estimation of model error and system state from measured data. We will design optimal strategies for experimental planning to resolve ambiguities of inferred model error and investigate the effect of model error on model reduction techniques. A long term goal is to devise computational strategies to automatically extend mechanistic models by black-box representations of relevant exosystem dynamics not in the focus of the modellers interest. We aim to make the algorithms and methods available as a free software package.
» weiterlesen» einklappen