Skip to content

README.md

Latest commit

 

History

History
102 lines (57 loc) · 3.56 KB

README.md

File metadata and controls

102 lines (57 loc) · 3.56 KB

📈 CS Student Strength Analysis

This project explores the relationship between academic performance and problem-solving ability among Computer Science students. Using data visualizations and statistical insights, we break down how performance in core CS courses correlates with the number of coding problems solved — a key indicator of applied skill.

🧠 Objective

To analyze whether higher academic performance (GPA) in technical subjects such as OOP, Algorithms, and AI correlates with problem-solving strength, and to identify which subjects show the most variance or consistency among students.

📁 Dataset

  • CS Student Strength.csv
  • Includes GPA data from 20+ CS courses (theory and sessional)
  • Also includes "Total Problems Solved", indicating applied coding skill

📊 Visualizations

1. 📊 Average GPA Per Course

Screenshot 2025-04-15 at 8 14 50 AM

Bar plot showing mean GPA across all subjects
✅ Conclusion: Students performed best in DBMS, Artificial Intelligence, and Discrete Math. Subjects like Machine Learning and Algorithm showed lower average GPAs.

2. 🧪 Grade Distribution in Core CS Subjects

Screenshot 2025-04-15 at 8 15 09 AM

Box plot of GPA distribution in key subjects
✅ Conclusion: Subjects like Data Structure, Operating Systems, and OOP had a wider GPA range, suggesting varying student comprehension. Compiler Design showed a tighter distribution.

3. 📈 Problem-Solving vs Course GPA

Screenshot 2025-04-15 at 8 15 41 AM Screenshot 2025-04-15 at 8 16 36 AM

Scatter plots comparing "Total Problems Solved" with GPA in:

  • OOP
  • Algorithm
  • Artificial Intelligence

✅ Conclusion: There appears to be a positive correlation — students with higher GPAs in these subjects tend to solve more problems. This supports the idea that academic performance and applied skill are linked, though not perfectly.

4. 🔥 Correlation Heatmap

Screenshot 2025-04-15 at 8 17 11 AM

Heatmap of correlations between all subjects and total problems solved
✅ Conclusion:

  • Moderate correlation between Artificial Intelligence, Algorithm, and problem-solving.
  • Some theory-heavy subjects (like Theory of Computing) showed weaker correlations, indicating that not all GPA scores predict practical ability.

📌 Key Insights

  • Students who perform well in AI, Algorithms, and OOP tend to solve more coding problems.
  • Some students excel at problem-solving even if they struggle in theory-heavy courses.
  • There’s a strong need to balance theoretical and applied learning in CS education.

🧰 Built With

  • Python 3
  • pandas, NumPy
  • matplotlib, seaborn

🧠 What I Learned

  • How to clean and analyze structured academic data
  • How to visualize relationships between variables
  • That GPA is useful, but not always the best predictor of applied CS strength

👨‍💻 Author

Emdya Permuy-Llovio