This repository contains a project focused on predicting student academic performance using machine learning techniques. Leveraging Linear Regression and multi-class features as descriptive variables, the project aims to build predictive models to forecast student grades. The dataset, sourced from Kaggle, includes 32 features and is used to train and evaluate various models, including a novel approach that incorporates target features as inputs for improved performance.
- Algorithm Used: Linear Regression
- Data: 612 instances with 9 attributes including grades, study time, absences, and parental education.
- Novel Approach: Utilizes previous grades as descriptive features to enhance prediction accuracy.
- Evaluation Metrics: Root Mean Squared Error (RMSE) and R² Score.