Research on Prediction of College Students' Registration Based on Machine Learning and Voting Model

Abstract

The registration rate of freshmen has always been a major concern for many colleges and universities, especially private ones. The author of this paper collects and completes a dataset of freshmen registration and uses various machine learning algorithms, including decision tree, random forest, and BP neural network, to learn from it. The author introduces the confusion matrix and F1 score to evaluate the effect of machine learning. A voting model based on multiple machine learning algorithms is designed to optimize prediction and the effectiveness of this scheme is verified through numerous experiments. The experimental results also reveal the factors affecting the registration of college freshmen, which has certain guiding significance for the enrollment of colleges and universities. 75% of the data is used as the training set and 25% as the test set. The data is preprocessed to make it standardized and complete. The results show that the performance of the voting model is significantly improved compared to a single algorithm, and the prediction accuracy of freshmen registration is maintained at more than 60%, with the F1 score reaching 0.7.