We revisit the protein secondary structure prediction problem using linear and backpropagation neural network architectures\r\ncommonly applied in the literature. In this context, neural network mappings are constructed between protein training set\r\nsequences and their assigned structure classes in order to analyze the class membership of test data and associated measures of\r\nsignificance. We present numerical results demonstrating that classifier performance measures can vary significantly depending\r\nupon the classifier architecture and the structure class encoding technique. Furthermore, an analytic formulation is introduced in\r\norder to substantiate the observed numerical data. Finally, we analyze and discuss the ability of the neural network to accurately\r\nmodel fundamental attributes of protein secondary structure.
Loading....