Font Size: a A A

The Research Of Data Mining Based On Tianjin University Of Finance And Statistics

Posted on:2017-02-28Degree:MasterType:Thesis
Country:ChinaCandidate:H ZhangFull Text:PDF
GTID:2348330515481385Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
Due to the recent development of information technology,the student-information management system is widely used in universities,which greatly improve the efficiency of school management.However,the system usually has some simple functions only,such as search or storage,which can not effectively utilize the hidden value of data.Our School has accumulated a large number of test data during these years.How to use the stored data in the system,and get useful information for school,has become the top priority.This article uses data-mining technology,through the analysis of the school’s statistics curriculum score in recent two years,hopefully can dig out valuable information,and provide helpful advices for the teacher’s teaching and students’ self-improvement.The article introduces the concept of data mining technology,focus on K-Means algorithm and Apriori algorithm,and applies these two algorithms in the process of score analysis.In this paper,the source of data is the mark of statistics curriculum taken by undergraduate students at Tianjin university of finance and economics in the past two years,on this basis,this article studies relevant score information.Preconditioning the raw data,deleting the missing items,and this article also generating four new indicators according to the knowledge point of relevant material:basic statistics,mathematical statistics,time series,and index analysis.At the same time,due to the need of association-rules analysis,the paper discretizing the data of the four indexes and the overall performance data of students,then acquiring a rough understanding through the descriptive statistics analysis of gender differences in score,teacher differences in score,type of questions and faculties differences in score.After the application of K-Means algorithm for data clustering,the different students are divided into five categories according to the result,each type of students are given different guidance.Furthennore,this paper applies the Apriori algorithm to take a further step to analysis the correlation rule among faculties,gender,overall grade,and grade of different subjects,thus finds the related rules among the factors,and puts forward suggestions to give a priority importance to basic statistics knowledge.
Keywords/Search Tags:student achievement, data mining, association, rules clustering
PDF Full Text Request
Related items