The aim of this project was to create a tool for improvement of microemulsions formulation process. Microemulsions are promising drug forms allowing for improvement of solubility and enabling controlled release of drugs. As a modeling tool, artificial neural networks (ANNs) were chosen. ME _expert system was created basing on the neural expert committee constructed of 11 neural models. ANN were designed to predict occurrence of microemulsion basing on its particular quantitative and quantitative compostion. Generalization abilities of ME _expert were tested in laboratory trials, where 77% of accuracy was found for unknown composition of microemulsion.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW