Comparison of Classification Data Mining in Process Majors Students
暂无分享,去创建一个
The process of admission of new students, especially in vocational school is the first step of school in selecting the prospective students in accordance with the chosen majors. But sometimes the selection of majors becomes boomerang for the students because after received in the department, they fail in the process of studying such as not match with the chosen majors when he has obtained the lessons, students learning interest is reduced so that the values of academic achievement decrease and if the student graduate, the student will find difficulty in finding a job that suits for their interests, even to the drop out. This is certainly a major problem for school management in maintaining these students and it impacts on the costs of education that have been issued, either by parents or by the government who subsidize the school to be useless, because the students do not have sufficient ability for the chosen majors. This study conducted a comparison of several methods of data mining classification in order to obtain the best accuracy value in the process of data training, testing and validation. This research can provide solution on the best classification method in the selection of majors, so it can be used as a reference in choosing the best method and suitable in similar cases, which will result in the students majors to be more accurate. So as to contribute to the school can make an appropriate decision to determine the students majors in accordance with the interests and talents of each student, thus minimizing students wrong in determining the majors that can result directly with the academic achievement of the student. The results obtained that KNN and ID3 method tops in accuracy of both data testing and validation with a value of 100%, C4.5 about 98.61% and 100%, Naïve Bayes 99.31% and 96.45% and the last Random Forest 92.36% and 89.92%. Software Quality Assurance obtained the value of 80. This means that the application is used to assist the management of majors.