IMPROVING EFFICIENT STUDENT CAREER RECOMMENDATIONS USING MACHINE LEARNING
Sumitra Nuanmeesri, Sumitra Nuanmeesri, Lap Poomhiran · Journal of Southwest Jiaotong University · 2023
This research aims to develop and improve the model’s efficacy for student career recommendations using feature selection, data sampling, and machine learning techniques. The results showed that the wrapper approach improved the model’s performance when applied to over-sampling techniques. In particular, a model combining k-mean, Synthetic Minority Over-sampling Technique, and Edited Nearest Neighbor methods will significantly improve model performance. Additionally, the model developed based on machine learning showed that the model using the Multi-Layer Perceptron Neural Network technique had a higher efficiency than other classifiers in this experiment. Finally, when developing the model using the wrapper, k-mean, Synthetic Minority Over-sampling Technique, Edited Nearest Neighbor, and Multi-Layer Perceptron Neural Network, the model is suitable predicts student careers. This model has a prediction accuracy of 95.88% and a root mean squared error of 2.0968. The most accurate models using the classifier are Support Vector Machine, k-Nearest Neighbors, Random Forest, Decision Tree, Naïve Bayes, and Logistic Regression. Keywords: Career Recommendations, Machine Learning, SMOTE-ENN, Resampling, Wrapper DOI： https://doi.org/10.35741/issn.0258-2724.58.2.49