Identifying and Training Data for Appraoching AI by Construction of Entropy and PSO

Abstract

In the current field of AI, we use neural network and classification and clustering of data mining that those sciences train data for building model.  However, those training data method still need domain knowledge of humans to sort relative data.  So, we provide a construction to similar creature learning method. In physical space and in the dynamic environment the data seem chaos, but they do have regulars that we can find. Furthermore, our intelligence is dependent on our abilities of identifying the relation between outside environment and us, and we need to recognize the environmental data that this recognizing behavior or process is like learning system via  logic and sorting functions.  When each event  is happened that we need to identify what data is affected by them.  If we want to directly distinguish those relative data that we use entropy to analyse the dynamic data of environment with social function of particle swarm optimization(PSO) for achieving the learning method of creature.