The Application of KNN Algorithm Based on Time Factor in the Warning Mechanism of Dismissal in Colleges and Universities

Many colleges and universities have employed a mechanism to dismiss students based on their academic status. However, traditional criterions only evaluate students' scores and credits without being conscious of the causes of poor academic performance, such as poor psychological health. In order to promote educational equality, we propose a method that evaluates students' psychosomatic well-being based on their behaviors. Our method will offer educators an opportunity to provide timely counseling services to students who bear psychological sufferings. First of all, we examined the possible indicators of students' psychological conditions and selected some features by considering behaviors and psychology. Then, we used the improved KNN algorithm combining with factors of months to predict which student will fail to complete their studies in the universities based on historical data. From the above analysis, we aim at creating a system that monitors the students' behaviors and detects their negative emotions. The system will help the universities to cope with students who are struggling in a timely manner and create a healthier campus environment.