The Study Of Extended Rough Set Model And Its Application In Supplier Selection

Rough set theory is another mathematical tool to deal with uncertain data besides probabilitytheory and fuzzy set theory. It is proposed by Pawlak in1982. The character of the theory is that itdoes not require any prior knowledge or additional information. It has a good application prospect inindex selections and program options.Classical rough set focus on complete information system, but we are often faced withincomplete information to access knowledge due to data measure error, data access restrictions andother reasons in real life. In other words, some attribute values of the objects are unknown. Based onfuzzy decision variable precision rough set model under complete information, this thesis proposedrough set attribute reduction to deal with multi-attribute decision making problems under incompleteinformation. A numerical example is used to test the feasibility of the model.In addition, sorting optimal selection is another problem in multi-attribute decision makingproblems. Most previous studies default that the attribute values are compensate, but there also exitsome situations that attribute values are not compensate. This thesis also proposed an extended roughset model to sort optimal selections. It uses rough set conditional information entropy to get theweights. Combining calculating the relative information entroy to considerate the integrity andbalance of each object, we can improve the accuracy of the results under a centain degree.Finally, the sorting model is applied to the selection of suppliers. We construct the index systemfor chemical equipment spare part’s supplier selection and collect the information of the suppliers.Then the sorting model is applied to indicate its value. The calculation results show that the improvedmodel is more in line with the enterprise’s actual choice. So the method is scientific and effective.