Data Mining research of bachelor's degree dissertation of College of Electronic and electrical engineering in a university in China

This study tries to find out some rules and characteristics through the data mining of students' dissertation topics, such as the College of Electronic and Electrical Engineering, dissertation of undergraduate students majoring in Electrical Engineering and Automation in the past three years. The study employs the quantitative method of research. In this study, we will use the process of K-DD(Knowledge Discovery in Database). Using logical regression, it is to predict whether the student can reach the average level of students. Linear multiple regression was used to study the final score of students' graduation project depending on what factors. Through the method of decision tree, according to the basic level of each student and the subjects that he is good at, determine the direction of the paper topic that the student is suitable for. The conclusion of this paper shows that the main topics of the dissertation of Electrical Engineering and Automation are control, system design and simulation. The main software used is matlab. The results of logical regression and linear regression show that there is a sexual relationship between the final score of the thesis and the evaluation and defense performance of fellow teachers and instructors, and the final decision tree can also provide constructive guidance for the selection of the topic of the thesis.