Student Performance Analytics System

A Python-based data analysis project that explores the relationship between student engagement (attendance, library visits, LMS logins, lab usage, assignment submissions) and academic outcomes (midterm and final marks). The project cleans a raw dataset, engineers a composite performance score, runs statistical hypothesis tests, and visualizes the results.

Features

• Cleans raw student records by handling missing values and removing duplicate entries
• Generates descriptive statistics (mean, min, max, std) for all numeric columns
• Groups and compares performance by department and by semester
• Identifies top-performing students and flags students who may need academic support
• Answers targeted analytical questions (e.g. which department scores highest, which factor correlates most with final marks)
• Normalizes engagement-related features to a 0–100 scale using NumPy
• Builds an Engagement Score from weighted normalized features using matrix multiplication
• Computes an Academic Score and a final weighted Performance Score for every student
• Runs two hypothesis tests with SciPy:
  • Pearson correlation test — attendance percentage vs. final marks
  • Independent samples t-test — final marks for high vs. low LMS login frequency groups
• Visualizes attendance distribution, department-wise performance, semester-wise averages, the attendance–final marks relationship, and the breakdown of students into performance categories
• Categorizes students into Top Performer, Average, and At Risk groups
• Exports the fully processed dataset to a new CSV file

Technologies Used

• Python 3

Libraries Used

• pandas — data loading, cleaning, and aggregation
• numpy — normalization and matrix-based score calculations
• scipy — hypothesis testing (Pearson correlation, t-test)
• matplotlib — data visualization

Folder Structure

Student-Performance-Analytics-System/
│── student_performance_analysis.ipynb
│── student_data.csv
│── student_data_processed.csv
│── README.md
│── requirements.txt
│── LICENSE
│── .gitignore

Installation

1. Clone the repository:

git clone https://github.com/<your-username>/Student-Performance-Analytics-System.git
cd Student-Performance-Analytics-System

2. Install the required libraries:

pip install -r requirements.txt

How to Run

1. Open the notebook in Jupyter Notebook, JupyterLab, or Google Colab:

jupyter notebook student_performance_analysis.ipynb

2. Run the cells in order from top to bottom.
3. The notebook generates student_data.csv on first run and saves the cleaned, feature-engineered dataset to student_data_processed.csv at the end.

Dataset Information

The dataset contains 20 unique student records (plus one intentionally duplicated row to demonstrate duplicate handling) across four departments — CS, AI, SE, and DS. Each record includes:

• Student ID, Department, Semester
• Attendance Percentage
• Library Visits per Month
• LMS Login Frequency
• Lab Usage Hours
• Assignment Submission Rate
• Midterm Marks and Final Marks

The dataset is generated directly inside the notebook and also included as a standalone CSV file for reference. It contains two intentionally missing values (Library Visits and Lab Usage Hours) to demonstrate the data-cleaning workflow.

Analysis Performed

• Missing value detection and imputation using column means
• Duplicate detection and removal
• Descriptive statistics across all numeric features
• Department-wise and semester-wise performance summaries
• Top 5 performer ranking and at-risk student identification (attendance below 70% or final marks below 50)
• Correlation analysis to find which engagement factor relates most strongly to final marks
• Feature normalization (min-max scaling to 0–100)
• Weighted Engagement Score calculation using matrix multiplication
• Academic Score and final Performance Score calculation
• Pearson correlation test: attendance percentage vs. final marks
• Independent samples t-test: final marks by LMS login frequency (high vs. low groups)

Visualizations Generated

• Attendance distribution across students
• Department-wise performance comparison
• Semester-wise average marks
• Attendance vs. final marks scatter relationship
• Performance category breakdown (pie chart)

Skills Demonstrated

• Data cleaning and preprocessing with pandas
• Feature engineering and normalization with NumPy
• Statistical hypothesis testing with SciPy
• Data visualization with Matplotlib
• Exploratory data analysis and reporting on a structured dataset

Future Improvements

• Expand the dataset with more student records for more robust statistical results
• Add predictive modeling (e.g. regression) to estimate final marks from engagement features
• Turn the analysis into a reusable script or simple dashboard
• Add automated tests for the data-cleaning and scoring functions

Author

Farhan Khizar

License

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