Research Of Decision Support System For Agriculture Based On Rough Set

Rough set theory is a new mathematical tool to deal with the imprecise, incomplete and uncertain knowledge. In current information explosion era, the powerful functions of DSS on assistant decision-making and rough set on information processing is becoming increasingly obviously, and have called people's attention. According to the characteristics of the area of agriculture data, this thesis proposed the model of decision support system(DSS) for agriculture application based on rough set theory.This thesis discussed the current situation and characteristics of DSS and rough set theory, and put forward the solution of DSS for agricultural based on rough set theory. This method discussed from the decision table of knowledge acquisition, and made use of rough set theory to reduce the table in order to provide rule table for DSS to make assistant decision-making. This thesis mainly discussed the method of the attribute reduction of rough set. Among the existing research results, the discernibility matrix proposed by Skowron provided an outstanding thread for acquiring the most ideal knowledge reduction. Since the discernibility matrix-based attribute reduction method involves the combination of attributes and transforming it into a matrix, and can find the core of attributes easily. Therefore this thesis studied the important role of discernibility matrix in finding the core of attributes, then proposed an improved attribute reduction algorithm for incomplete information systems based on the limited tolerance relation. According to the task of knowledge acquisition in DSS,we built the model of DSS based on rough set theory. In this model, we firstly analyzed the pretreatment method for agricultural data, which included the discretization and generalization issues of attributes. We pointed out that the best method is to combine with expert's experience. After pretreatment, the data will form a target information system, then attribute reduction algorithm is used to get rules, then achieved the purpose of application of rough set in agricultural DSS. Last, the rules will be used in the diagnosis of pests and diseases, and the method of importance of attribute can give credibility to the diagnosis. Through the comparison and analysis to the example, experimental results show that the plan of agriculture DSS based rough set is feasible and the algorithm of attribute reduction is effective and practical.