Granular Neural Networks with Decision Fusion

A granular neural network is designed to deal with granular data fusion and granular knowledge discovery. From a data granulation point of view, the granular neural network can process granular data in a database. From a knowledge discovery point of view, the granular neural network is able to learn internal granular relations between inputs and outputs, and predict new relations. From a decision fusion point of view, the granular neural network is capable of merge multiple granular decisions to generate a robust decision. In general, the granular neural network is a systematical network with relevant intelligent components dealing with data granules, knowledge granules and decision granules.