Abstract: Fuzzy min max (FMM) model is a combination of both fuzzy set and neural network for classification. It uses hyperbox structure for pattern classification which consists of min and max points of opposite corners of hyperbox. To handle overlapping region of hyperboxes during classification is a crucial role. There are various FMM models are described in which during learning phase the hyperboxes are expanded until almost the whole pattern space is covered. At the end of learning phase, there are no overlapping hyperboxes that belong to different classes. The maximum expansion of these hyperboxes is controlled by the expansion parameter ? which is used for expansion.

Keywords: Fuzzy min-max (FMM) model, hyperbox, neural network, pattern Classification.