Analysis For Pre-warning Students Based On DM And Its Application

In recent years, higher education is quickly developing with the rapid advance of science and technology. Colleges and universities continue expanding the enrollment. The number of students has constantly increased. Simultaneously, the quality of students generally becomes worse. These bring some problems in higher education and teaching process. In the daily teaching, colleges and universities have accumulated massive and complex data that is more difficult to manage and analyze these data. It is very useful to discover meaningful and valuable information from these data and the analysis results are helpful for decision-making, supervising the process of teaching, improving college management.It is an inevitable trend that data mining is applied in college teaching and management since the application of data mining becomes wide. Data mining is used to fully analyze hidden and internal relation between examination results and various factors in order to take effective measures, advance teaching reform and increase teaching quality.This thesis is based on the given pre-warning data and mines the given data in accordance with the corresponding analytical patterns through data mining technique (classification and prediction, association analysis and clustering analysis).By means of data mining, the potential relations among the courses that students easily fail can be obtained , the possible causes and key factors that caused the failure can be discovered, thus the characteristics and commonness of the pre-warning students can be gained. The analysis results help discovering the hidden rules and patterns. This information is helpful for avoiding a student to become a pre-warning one, predicting the teaching work, ensuring the whole teaching quality.