** project overview **

Student Performance-BD: Analyzing Academic Outcomes Using EDA & Hybrid Models

Overview

This project aims to analyze and predict the academic performance of students from various regions in Bangladesh. Using the Student Performance-BD dataset, which includes 24 attributes related to demographics, academic metrics, extracurricular activities, and other factors, we perform Exploratory Data Analysis (EDA) and develop a Hybrid Machine Learning Model to predict students' academic success.

The insights from this project can be used to identify key factors influencing academic outcomes and help guide students in choosing the most suitable educational path based on their individual profiles.

---

Dataset

The Student Performance-BD dataset contains information about students' demographics, academic performance, extracurricular participation, and other relevant factors. Key attributes include:

• Demographics: Age, Gender, Residence (Urban/Rural)
• Educational Background: Academic level, Group (Science/Commerce), etc.
• Academic Scores: English, Math, Science, Social Science, etc.
• Extracurricular Activities: Participation in activities like art/culture.
• Other Factors: Attendance, Type of school (Private/Govt/Semi-Govt), etc.

Dataset Sample

Student ID	Name	Age	Gender	Residence	Group	Attendance	English	Math	Science	Social Science	Art/Culture	Academic Scores
0	Avi Biswas	16	Male	Urban	Science	95	95	98	92	94	Yes	98
1	Taslima Sultana	18	Female	Rural	Commerce	92	65	71	40	78	No	80
2	Md Adilur Rahman	15	Male	Rural	Commerce	81	64	78	58	86	Yes	74
3	Saleh Ahmed	16	Male	Rural	Science	90	84	90	85	86	Yes	88
4	Din Islam	17	Male	Urban	Commerce	75	54	70	45	79	Yes	76

---

Project Goals

• Perform Exploratory Data Analysis (EDA) to uncover patterns and correlations in the dataset.
• Preprocess the data for machine learning model training.
• Develop a Hybrid Machine Learning Model combining multiple algorithms (e.g., Random Forest, Gradient Boosting).
• Predict student academic performance across various subjects.
• Evaluate and visualize model performance using metrics such as accuracy, precision, recall, and ROC-AUC.

---

Project Structure

Student_Performance_BD/
├── data/
│   └── student_performance_bd.csv        # Dataset file
├── notebooks/
│   └── EDA_and_Model_Development.ipynb   # Jupyter notebook with EDA and modeling steps
├── src/
│   ├── preprocessing.py                 # Data cleaning and preprocessing scripts
│   ├── models.py                        # Machine learning model implementation
│   └── evaluation.py                    # Model evaluation and performance visualization
├── results/
│   └── plots/                           # Folder for generated plots
├── README.md                            # Project documentation
├── requirements.txt                     # Python dependencies
└── LICENSE                              # Project license

---

Installation

Follow the instructions below to set up this project on your local machine.

1. Clone the repository:

   git clone https://github.com/your-username/Student_Performance_BD.git

2. Navigate to the project directory:

   cd Student_Performance_BD

3. Install dependencies: Use pip to install the required packages listed in requirements.txt:

   pip install -r requirements.txt

---

Usage

Run the EDA and Modeling Notebook:

jupyter notebook notebooks/EDA_and_Model_Development.ipynb

Data Preprocessing and Model Training:

Run the preprocessing and model training scripts in Python:

python src/preprocessing.py
python src/models.py

Evaluate the Model and Visualize Results:

Use the evaluation script to visualize model performance:

python src/evaluation.py

---

Features

• Exploratory Data Analysis (EDA):

  • Statistical analysis (mean, median, correlation).
  • Visualizations: Histograms, Box Plots, Pair Plots, Heatmaps.

• Machine Learning Models:

  • Hybrid model combining Random Forest and Gradient Boosting for predicting student performance.
  • Model hyperparameter tuning and cross-validation.

• Model Evaluation:

  • Accuracy, Precision, Recall, F1-Score, and ROC-AUC curves.
  • Learning Curves and Confusion Matrices for evaluating model performance.

---

Results

The hybrid model achieved the following performance metrics on the test data:

• Accuracy: 0.9640162507254788
• Precision: 0.9505736244243131
• Recall: 0.8347133112725867%
• F1-Score: 0.8813750794053076
• ROC-AUC: 0.9943288855613831

Key Insights:

• Students' attendance, extracurricular activity participation, and group choice (Science/Commerce) were significant predictors of academic performance.
• Urban students tend to have better scores in subjects like English and Math compared to rural students.

---

Contributing

Contributions are welcome! If you’d like to contribute, follow these steps:

1. Fork the repository.
2. Create a new feature branch:

   git checkout -b feature-name

3. Commit your changes:

   git commit -m "Add feature-name"
   git push origin feature-name

4. Open a pull request.

---

License

This project is licensed under the MIT License. See the LICENSE file for more information.

---

Contact

Author: Arif Mia
📧 [arifmiahcse@gmail.com]
💼 [www.linkedin.com/in/arif-miah-8751bb217]

Feel free to reach out for collaborations or feedback!