Classification of cancer recurrence with alpha-beta BAM (Q1036428)

Summary: Bidirectional Associative Memories (BAMs) (based on a model proposed by Kosko) do not have perfect recall of the training set, and their algorithm must iterate until it reaches a stable state. We use a model of Alpha-Beta BAM to classify automatically cancer recurrence in female patients with a previous breast cancer surgery. Alpha-Beta BAM presents perfect recall of all training patterns and has a one-shot algorithm; these advantages make Alpha-Beta BAM a suitable tool for classification. We use data from the Haberman database, and a leave-one-out algorithm was applied to analyze the performance of our model as a classifier. We obtain a percentage of classification of 99.98\%.