Research Of The Rough Set Model Based On Fuzzy Partial-order Relation And Data Mining Method

As a unique tool to deal with the problem of uncertainty,rough set theory has been extensively used in data mining,pattern recognition,machine learning,and other fields.At present,scholars tend to study the uncertainty measurement,attribute reduction,and rule acquisition of rough sets and all.Although some research achievements have been made in these aspects,it is still immature and needs to be further examined and explored.Firstly,a fuzzy partial-order relation rough set model is proposed in interval-valued continuous decision systems,and its properties and characteristics are discussed in detail.At the same time,the proposed model is simulated and calculated with the other rough set models.From the theoretical analysis and data simulation results,we can conclude that the new model proposed is more universal and reasonable.Secondly,we put forward a new method for measuring uncertainty based on the new model,and it can be used to measure the uncertainty of both knowledge and concept in the system,and the method is more accurate compared to the other methods.Therefore,it lays a foundation for accurate attribute reduction and knowledge classification.Finally,an attribute reduction algorithm based on fuzzy rough entropy is proposed in interval-valued continuous decision systems,and the time complexity of the algorithm is discussed.Simultaneously,how to extract and optimize the rules in the reduced system is discussed,and the "precision" and "strength" of the decision rules can be ensured by setting coverage and accuracy.In the end,a comprehensive algorithm of rule extraction and optimization is constructed,which lays a feasible foundation for the development of the practical application system.