Attribution Reduction of Battlefield Electromagnetic Measurement Based on Fuzzy-Rough Information Entropy

To study the battlefield electromagnetic measurement data effectively, it is very necessary to research a valid attribute reduction method for dealing with the measurement information. Firstly, the deficiencies of some current information entropy models are analyzed, and a new kind of generalized information entropy model based on fuzzy-rough set is proposed, which can introduce the probability distribution of objects in the universe into the entropy calculation. Secondly, a measurement of significance of attributes based on the proposed model is studied, and then a heuristic fuzzy-rough data reduction method based on the information entropy has been put forward. Finally, four teams of measurement data of electromagnetic target in a certain battlefield have been chosen for constructing the fuzzy decision table, and the method discussed above is applied on the example for attributes reduction. The calculation result has shown the validity of attribution reduction of battlefield electromagnetic measurement based on the proposed method.