Analysis Of Students’ Abnormal Behavior In Fuzzy Rough Set Based On Hierarchical Structure

With the development of smart campus and big data,more and more technologies combine data mining with school education to mine student behaviors and information conducive to better campus management,timely discover students’ abnormal behaviors and report them to relevant personnel to ensure better education,management,practice and other activities of the school.In the study of students’ abnormal behavior analysis,there are few studies considering the hierarchical relationship and the dynamic change of students’ data.Therefore,this paper proposed the fuzzy rough set algorithm based on hierarchical structure,and puts forward the fuzzy rough set incremental algorithm based on hierarchical structure under the change of objects and attributes for the analysis of students’ behavior.The main research contents are as follows:1.The fuzzy rough set non incremental algorithm based on hierarchical structure is proposed.The upper approximation of non-leaf node is obtained by its child node,the lower approximation of non-leaf node is determined by its sibling node,and the sibling node without sibling node is found to be the closest ancestor node.On this basis,the abnormal behavior of students is analyzed.The fuzzy rough set non incremental algorithm based on hierarchical structure is used to classify the students’ behavior.Compared with the original classification results,a class of students with no discipline category is added,which is generated by non-leaf nodes,and the ranking of students under different attributes is analyzed for the new class of students’ abnormal behaviors.2.For the dynamic changing data,when the object changes,the fuzzy rough set increment algorithm based on hierarchical structure is proposed,and the proof and examples are given.The fuzzy rough set increment algorithm based on hierarchical structure is used to analyze the characteristic data of students’ campus,find out the number of students’ abnormal behavior,and explain the characteristics of abnormal behavior.Finally,through the experiment of student campus data over 3000 and student card consumption data over 60000,the running time of fuzzy rough set algorithm,the non-incremental algorithm based on hierarchical structure fuzzy rough set and the incremental algorithm based on hierarchical structure are compared.The experimental results show that the incremental algorithm based on hierarchical structure is suitable for the data set with large amount of data,multiple categories and high frequency of object change.3.In the dynamic changing data,when the attributes change,from the fuzzy similarity relation,the fuzzy rough set increment algorithm based on hierarchical structure is proposed when the attributes increase and attributes decrease,and the corresponding proof and algorithm are given.The incremental algorithm of fuzzy rough set based on hierarchical structure is used to find out the students with abnormal behaviors,and the characteristics of abnormal behaviors are described.Finally,when the attributes change,experiments are carried out on the characteristic data of student campus over 3000 and the consumption data of student card over 60000,and the fuzzy rough set algorithm based on hierarchical structure is compared and analyzed.The experimental results show that the fuzzy rough set incremental algorithm based on hierarchical structure is suitable for the data sets with large amount of data,multiple categories and high frequency of attribute change.