Global E-commerce Customers Behavior Dataset

This dataset contains detailed e-commerce transaction records representing customer demographics, purchase behavior, and shopping preferences across multiple countries. It is designed for exploratory data analysis, customer segmentation, and machine learning tasks such as predicting returning customers. The dataset closely reflects real-world online retail behavior.

🔑 Key Highlights

• Over 100,000 customer transactions
• Includes demographic, transactional, and behavioral features
• Suitable for EDA, classification, and imbalance-handling techniques
• Ideal for machine learning and data science portfolios
• Clean, structured, and Kaggle-ready format

---

📊 Detailed Description

Each row in this dataset represents a single transaction from a global e-commerce platform. Customer information such as age, gender, location, and device used is combined with purchase details including product category, payment method, and transaction amount.

The target variable, ReturningCustomer, indicates whether a customer made repeat purchases. This makes the dataset particularly useful for customer retention analysis and predictive modeling. Due to natural class imbalance, the dataset is also suitable for applying advanced techniques such as SMOTE, class weighting, and ensemble-based machine learning models like Random Forest and XGBoost.

This dataset can be used for:

• Customer behavior analysis
• Retention and churn prediction
• Marketing strategy optimization
• Machine learning experimentation and portfolio projects