Rough Sets And Its Applications In Evaluation For Cigarette Quality

Rough Sets theory is a new mathematical tool of dealing with imprecise and uncertain information.It has been applied to many areas including machine learning,data mining and intelligent data analysis.The thesis starts with the basic theory of rough sets and ends with the application of rough sets in information processing,from theory model to algorithm and practical application system,which contains the main research results of the author's during postgraduate.Recent researches and developments on rough sets theory are described in first chapter,and the basis of rough sets theory is introduced in second one,which including knowledge and classification,decision table,the basic concept of rough sets and knowledge acquisition based on rough sets theory.In chapter three, a new method of roughness measurement to rough set is defined based on equivalence relation and general binary relation.The concept of boundary entropy is presented which denoted by knowledge roughness of rough set boundary,together with rough degree,a more reasonable method of measuring the roughness of rough set is introduced which provides theoretic basis for concept acquisition and description in rough set theory.In chapter four,variable precision rough set model is extended based on general binary relation.Variable precision rough rules set,reliability and importance of rules are discussed.At last,an algorithm of rule mining is given based on general binary relation.All this provides theory basis for more extensive applications of variable precision rough set model. In chapter five,the shortcoming of traditional operator of rough fuzzy sets is analysed,then a new approximation operator of rough fuzzy sets in accordance with the drawback is probed.It elaborates that the new operator is reasonable in dealing with information and explains that the new method is efficient by means of an...