This system is developed to predict the re-enrollment of prospective students at Universitas Selamat Sri Batang using the Random Forest algorithm. The model helps university administrators identify students who are at risk of not completing their re-enrollment process.
- Develop a classification model to predict student re-enrollment.
- Analyze the factors influencing student re-enrollment.
- Python: The primary programming language.
- Flask: A web framework for displaying prediction results.
- Random Forest: The main algorithm for classification.
- SMOTE: To handle data imbalance.
- Chi-Square: For feature selection.
- Scikit-learn: For machine learning implementation.
- Pandas & NumPy: For data manipulation.
- Re-Enrollment Prediction: The model can predict whether a prospective student will complete re-enrollment..
- Preprocessing Data: Cleaning and transforming data before modeling.
- Data Balancing: Using SMOTE to address data imbalance.
- Feature Selection: Using Chi-Square to select the most influential features.
- Model Evaluation: Utilizing confusion matrix, accuracy, precision, and recall.
The dataset consists of admission records of prospective students at Universitas Selamat Sri Batang from the 2020 to 2023 academic years, totaling 1,358 records, with 12 attributes and 1 target class.
- Model Accuracy: 93.76% after balancing with SMOTE and feature selection using Chi-Square.
- Confusion Matrix: Used to assess model predictions.
- Evaluation Metrics: Precision, Recall, and F1-Score.
Ahmad Muyaqi
Computer and Information Technology Education
Universitas Negeri Semarang
For any inquiries or suggestions, please contact LinkedIn https://www.linkedin.com/in/ahmadmuyaqi/.