A novel neural network model for different information granulation processing

Abstract

This paper presents a neural network model for processing different information granulation. Based on the requirement for processing mass information, different hierarchies of the information granulation are constructed. The basic concepts and researches of information granulation are introduced in this paper. And the model that can accept different inputs of information granulation in different levels is presented. Also this paper offers the deduction of the supervised learning algorithm for the neural network taking the three levels of granulation for instance. An example of college student comprehensive estimation is stimulated by the presented model, and experiment results illustrate that the model is efficient.