l-isaro/PredictED-model-training-and-evaluation

# PredictED-Rwanda ## Overview PredictED-Rwanda addresses the challenge of identifying students at academic risk in Rwanda's secondary education system. Traditional assessments often miss early warning signs, delaying interventions. This project uses a machine learning model to predict academic failure based on the UCI Student Performance Dataset. The goal is to enhance model performance with optimization techniques. - **Dataset**: UCI Student Performance Dataset (student-mat.csv) with features like study time, absences, and final grade (G3 ` 2. Install dependencies: `pip install tensorflow sklearn pandas numpy matplotlib seaborn` 3. Place `student-mat.csv` in the project directory. 4. Open `summative-final.ipynb` in Jupyter Notebook or Google Colab and run all cells. ## Files - `summative-final.ipynb`: The main notebook with code and results. - `student-mat.csv`: The dataset. - `saved_models/`: Directory for saved model files (e.g., nn_instance4.keras). - `video_presentation.mp4`: youtu.be