Neural Netwroks/AI 2003 Readers’ Choice Finalist: NeuroShell Classifier
NeuroShell Classifier solves classification and categorization problems based on patterns learned from historical data. It produces outputs that are the probabilities of the input pattern belonging to each of several categories, such as acidic/neutral/alkaline or cancer/benign. The classification algorithms are a neural network and a statistical classifier driven by a genetic algorithm and have been optimized to solve classification problems. Statistical tools such as an agreement matrix (sensitivity and specificity), probability graphs, ROC curves, and input rankings assist in analyzing the effectiveness of the model.