Utilizing Genetic Algorithms To Enhance Student Graduation Prediction With Neural Networks

Abstract

The prediction of student graduation plays a crucial role in improving higher education efficiency and as-sisting students in graduating on time. Neural networks have been used for predicting student graduation; however, the performance of neural network models can still be enhanced to make predictions more accurate. Genetic algorithms are optimization methods used to improve the performance of neural network models by optimizing their parameters. The problem at hand is the suboptimal performance of neural networks in predict-ing student graduation. Thus, the objective is to leverage genetic algorithms to improve the accuracy of stu-dent graduation predictions, measure the improvements obtained, and compare the accuracy results between the genetic algorithm-optimized neural network model and the neural network model without optimization. The training process of the neural network model is conducted using training data obtained through experiments, and the accuracy results of the neural network model with and without genetic algorithm optimization are compared. The research findings indicate that by harnessing genetic algorithms to optimize the parameters of the neural network model, the accuracy of student graduation predictions increased by 2.78%. Furthermore, the Area Under the Curve (AUC) also improved by 0.037%. These results demonstrate that integrating genetic algorithms into the neural network model can significantly enhance prediction performance. Thus, this study successfully utilized genetic algorithms to improve student graduation predictions using a neural network model. Experimental results show that prediction accuracy and AUC values significantly increased after opti-mizing the neural network model's parameters with genetic algorithms. Therefore, the use of genetic algorithms can be considered an effective approach to improving student graduation predictions, thereby assisting educa-tional institutions in improving efficiency and helping students graduate on tim.