Research Of Uncertainty Reasoning Rules Mining Based On Pan-right Field Methodology

Rough Set is an important subject in data mining, which has an irreplaceable role in dealing with uncertain information. However, in many cases the rules we obtained by Rough Sets are not perfect. For example, it can not show the effectiveness of the rules, and its reduction function will likely get rid of the attribute we need from the database. For these reasons, the uses of Rough Set are weaked greatly. So we put Rough Set theory and uncertainty reasoning together to study on the way of Pan-right field theory, and study the uncertainty reasoning rules mining based on Pan-right field theory.This paper introduces the background and significance of research, then introduces the research status at home and abroad, and crystallizes the main research contents, and builds the construction, and shows the new idea. Through the general train of thought and the key issues to be resolved and the relative theoretical knowledge, we give a few aspects behind:Firstly, based on the knowledge classification in Rough Set, the paper develops the method and theory, and shows the selective reasoning to obtain the selective reasoning rules that is bias rules; meanwhile, it can gain the global rules based on the bias rule mining. So the mining method is more rational.Then, we develop the reduction rules in Rough Set, and analysis the reduction process by the Pan-right field theory, and put the reduction and the association rules together, and put the reduction and compatibility analysis together, so we gain the more effective (m, n) reduction rules.Finally, we put the methods into the military exercise information mining during the entire process of knowledge discovery, to guide the theory and the instruction method for the military exercise evaluation, to raise the combat effectiveness of great theoretical and practical significance.