Reserch On Hobbies Classification Of University For Data Mining

In recent years,many colleges and universities actively promote the construction of the school for individual development and construction.They found that the diversity and complexity of the interest will affect the school’s decision making The classification method can effectively reduce the complexity and diversity of the interest,when the school will be the student’s interest hobby as an important reference factor for in-depth study of the time.Machine learning can build a classification model with good generalization performance and precision for Small sample data set for high dimension.The method overcomes the problem that the traditional mathematical statistic is affected by the change of data quantity,the difficulty of data statistics,and the power consumption of the data.College Students’ daily life data,the structure of a single,higher degree of freedom.In this paper,based on the characteristics of the students’ attributes,the classification model is trained by the support vector machine(SVM),which is based on the interest category.The comparison of decision tree and BP neural network classifier is used.According to the experimental verification of 5 data sets,the test was repeated 6 times to calculate the average value,respectively to explore the students’ interest in sports,travel,entertainment,hobbies diet hobby hobby,classification elective preference,experiments show that the SVM classifier can dig in the premise of quality guarantee under the fast mining students’ interest categories,the classification effect is obvious better than the decision tree and BP neural network.This method can promote the reform of the talent training mode,and provide the basis for the individualized construction of University.However,the current research only aimed at the classification of students’ interest in the data cleaning and optimization methods have yet to be further improved.