Dissertation Student Grade Analysis And Research Based On Bayesian Network

In recent years,data mining technology has been widely used in various industries such as agriculture,commerce,finance,and other industries.Its effectiveness has been increasingly apparent,but its application in the field of education is still not very mature,it is difficult to realize its value and powerful features.Education is a business that is important to the entire society.Students' academic performance has always been a matter of greatest concern to parents and teachers.Therefore,it is necessary to use data mining technology to further study education issues.On the basis of consulting a large number of relevant literature and famous scholars'research,this paper 's analysis and research on the academic performance of fifth grade primary school students in Wuhan City.The main work includes the following aspects:(1)First,comprehensive analysis of factors that affect student achievement,then develop a questionnaire;Secondly,collect the data for the first semester of the 2016-2017 school year for all students in the six classes of the fifth grade of Wuhan W Primary School.Secondly,collect the data for the first semester of the 2016-2017 school year for all students in the six classes of the fifth grade of Wuhan W Primary School.For example,the results of the midterm exams in Chinese,mathematics,English,gender,and questionnaires.(2)A principal component analysis was performed on the data that influenced the student's performance factors.Then ten main influence factors were obtained,and the cumulative contribution rate of the calculation results was 82.11%.In other words,these 10 factors can represent 82.11%of the original data information.They are:student sleep,student happiness index,physical health,etc.(3)Divide the above 10 influencing factors into three categories:parental influences,teacher influences,classmate influences.According to the analysis of the results,the time of rest is regarded as the representative of the total influence of the parents on the student's performance;The teacher's charisma is regarded as the representative of the teacher's total impact on the student's performance;and the student's enthusiasm for attending school is considered as the representative of the student's total impact on the student's performance.Then do a corresponding analysis on this:Students with different grades of different grades are affected by the teacher,parents,and classmates.For some results are shown in the under figure:The graph reflects the degree of affection of teachers of different gender of different grades of students.Through this graph,we can see:? Among the students with grade A,the male students' academic performance is particularly affected by the teacher of the class or they have little influence;but about girls,it is no information.? Among the students with grade B,the male students' academic performance is much affected by the teacher's influence;the female student's academic performance is much influenced by the teacher or somewhat;? Among students with Grade C grades,about male students,it does not see information;the girls' academic performance is somewhat affected by the teacher or no;? Among students with Grade D,the male students' academic performance is much affected by the teacher or no;but about girls,it is no information.As a result,the significant result is that boys' performance is greatly affected by teachers.(4)The average scores of students' Chinese,Mathematics,and English are divided into four levels:A,B,C and D.And using the above 10 factors to build a Bayesian network model.The prediction results have a certain degree of feasibility,and it also explains the reliability of the influencing factors from the analysis.Among them,?(X)P(x1,x2,…,x10 | X):the joint distribution of the ten factors that represent grade X and affecting grades,P(x1,x2-…,x10):denote the distribution of ten factors that affect performance.P(X|x1,x2,…,X10|:denotes the distribution of students' performance grades under the top ten factors.The model is as follows:P(X|x1,x2,…x10)=?(X)P(x1,x2…x10|X)/P(x1,x2,…x10)Forecast result chart:This article has done a lot of work on the basis of comprehensively influencing the academic achievement factors of students.The advantages of this article:analyzing the factors that affect student performance;making a principal component analysis to get the main influencing factors;doing the corresponding analysis to get the daily performance of students with different gender grades;forecasting the student achievement grade.However,the further study of this paper is:selection of student's data objectivity,feasibility of the selected Bayesian network prediction model.