Design And Implementation Of Analysis System Based On Kmeans Algorithm Student Behavior

In the 21 st century, after experiencing the third generation of the Industrial Revolution, IT has rapidly developed. In all walks of life, IT industries are gradually moving towards digitization, informationization and automatization. To embark on the trend of the information age, students work management systems have been introduced to many colleges, to facilitate student management for counselors, and the system reduces the school management workload of the staff to a certain extent. Based on analyzing student work management system in various institutions, I found that almost all of the systems do not have intelligence analysis function of the students’ behavior characteristics, and counselors can only inquire about students’ performance results through the systems. It can’t provide convenient and efficient working environment for counselors.In this thesis, This thesis decided to conduct in-depth analysis of the students’ behavior through data mining of students’ behavioral characteristics. Data mining methods is mainly using cluster analysis, and the students are divided into several best categories, a total student feature classification is based on the students’ common characteristics which is determined by their behavioral characteristics, so we can provide specialized category management for counselor. This thesis applied a improved Kmeans algorithm in this thesis. Make clear students’ common characteristics by their behavior features to realize the chara cteristic classification of students, thus providing classification management specialized fo r counselors.In this thesis, optimized Kmeans algorithm is used to caculate the best cluster ing coefficient- K. Cluster analysis is used to analyze students’ behavior characteristics thr ough main behavioral characteristics of students in school as the six dimensions of Kmean s clustering--- all sorts of performance, the scholarship amount, bonus from competitions,etc---and the coefficient matrix of dimension.Finally, the key of my research is to draw the prominent feature in the same class by analysing the results of several types of behavioral characteristics of students. This thesis hope that through this research This thesis can provide valuable research data for student’s behavioral analysis, and This thesis also hope to make my own effort for students classified and intelligence analysis.