🔍 Exploratory Data Analysis (EDA) — Credit Card Dataset

A comprehensive Exploratory Data Analysis (EDA) on a credit card customer dataset using Python. This project demonstrates fundamental data science techniques including data inspection, statistical analysis, visualization, and insight generation.

---

📊 Dataset

• Source: Bank Data Clustering — K-Means & PCA
• Records: 8,950 customers
• Features: 18 columns (credit limit, payments, balance, etc.)
• Format: CSV

---

📓 Notebook Contents

The notebook EdaPro.ipynb covers:

#	Section	Description
1	Setup & Imports	pandas, matplotlib, seaborn, scipy
2	Data Loading	Read CSV from GitHub raw URL
3	First Look	df.head() and column inspection
4	Data Structure	df.info() — types, non-null counts
5	Statistical Summary	df.describe().transpose()
6	Correlation Analysis	Heatmap + top 5 strongest correlations
7	Missing Value Analysis	Null counts and percentages
8	Distribution Analysis	Histograms with mean/median markers
9	Outlier Detection	Box plots + IQR outlier counts
10	Skewness & Kurtosis	Distribution shape statistics
11	Pairwise Relationships	Pair plot (sampled if needed)
12	Key Insights & Next Steps	Summary of findings

---

🚀 Getting Started

Prerequisites

pip install pandas matplotlib seaborn numpy scipy jupyter

Run the Notebook

git clone https://github.com/skmudassir-it/Machine-Learning.git
cd Machine-Learning
jupyter notebook EdaPro.ipynb

---

🛠️ Tech Stack

• Python 3.x — Core language
• pandas — Data manipulation
• matplotlib — Base plotting
• seaborn — Statistical visualizations
• scipy — Statistical functions (skew, kurtosis)
• Jupyter Notebook — Interactive computing environment

---

📈 Key Findings

• Correlation analysis reveals which features move together
• Distribution plots show skewness patterns across columns
• Box plots identify potential outliers for further investigation
• Pair plots visualize multi-dimensional relationships

Full insights are embedded throughout the notebook with clear markdown explanations.

---

🔜 Next Steps

• Apply clustering algorithms (K-Means, DBSCAN) to segment customers
• Use PCA for dimensionality reduction
• Build predictive models for credit risk or spending behavior
• Handle outliers with capping or transformations

---

📁 Repository Structure

Machine-Learning/
├── EdaPro.ipynb      # Main EDA notebook (24 cells)
└── README.md         # This file

---

🤝 Contributing

This is a personal learning project, but suggestions are welcome! Feel free to open an issue or fork the repo.

---

📝 License

MIT — feel free to use and modify.

---

👤 Author

Shaik Mudassir

• GitHub: @skmudassir-it
• Email: skmudassir.it@gmail.com

---

⭐ Star this repo if you find it useful!