Research On The Analysis Of Students' Grades Based On Decision Tree

With the continuous rise of data mining technology and the large increase in the number of college students,schools have introduced educational administration and student work management systems to store student information and improve the management level.How-ever,there is so much educational data that they just stay in the storage,query and simple statistical stage without any benefits.It is the starting point of this study to learn how to use the large amount of idle data to help dig out the underlying connections between teaching and learning,discover the existing problems in the daily teaching process with which we can improve teaching techniques to promote students' progress.Educational data mining is a feasible and effective method in solving the waste of educational information resources.The data mining classification methods will be strongly mentioned in this paper,where the K-means algorithm based on the initial clustering center optimization and outlier preprocessing along with the decision tree R-C4.5 algorithm will be introduced.The corresponding performance analysis model will be constructed to preprocess the stu-dents' performance data,with the hope to improve the accuracy and speed of data mining.Thus,we can help improve the follow-up teaching effects,and provide favorable scientific basis for teachers to improve the training mechanism and teaching quality.For the purpose of understanding college students' learning status and the need for improve-ment of the training plan,a method of students' performance based on decision tree algorithm will be designed to in this paper to analyze the scores of "Basics of computer engineering"for students in foreign language college.The K-means algorithm will be used to discretize the performance information to determine the cluster center,and the improved C4.5 algo-rithm will be used to mine and draw the decision tree according to the mined data,which will help analyze the importance and influence of each attribute on students' performance,give analysis to the teaching effect of the course,and propose improvement of the later teach-ing methods.By combining clustering and decision tree algorithm with the performance of computer foundation as experimental object,the data mining analysis can find out the main effects which perform badly in students' achievements,obtain valuable information that is helpful to the teaching,get the relevant analysis of the difficulty of exams and provide useful teaching suggestions for the teachers to improve their teaching quality.