Digital Network with MIN/MAX Nodes for Pattern Recognition

This research deals with a method of pattern recognition based on neural networks composed of MIN/MAX nodes. The general concepts of the MIN/MAX nodes and the neural networks are outlined. We introduce in this work advanced technique that tried to improved the performance of digital learning networks or (n_ tuple method) for recognition by using the principles of MIN/MAX nodes and gray scale images. From these principles we reached to a good level of enhancement in performance for this network.