Customer Segmentation

This project implements unsupervised machine learning for customer segmentation using K-Means clustering on retail customer data. The Jupyter notebook analyzes shopping behaviors to identify distinct customer groups based on spending patterns and demographics, enabling targeted marketing strategies.

Project Overview

The notebook processes customer datasets with pandas and scikit-learn to perform clustering analysis. It uses elbow method and silhouette scores to determine optimal clusters, visualizing segments through scatter plots and pairplots.

Core Analysis Workflow

The project follows a complete unsupervised learning pipeline for customer segmentation.

• File upload prompts for CSVs like 'Mall_Customers.csv', loading into DataFrames for initial inspection and EDA.
• Data preprocessing includes feature scaling with StandardScaler on numeric columns like age, annual income, and spending score.
• Clustering applies K-Means with elbow method (inertia plot) and silhouette analysis to find optimal k value, typically 5 clusters for standard mall customer data.

Key Visualizations

Multiple plots illustrate clustering results and customer distributions.

• Elbow method plot shows inertia vs. number of clusters (k=1 to 10) to identify optimal segmentation point.
• Silhouette score plots validate cluster quality and separation.
• Scatter plots visualize clusters in 2D space (Annual Income vs. Spending Score), color-coded by cluster labels.
• Pairplots and heatmaps reveal feature correlations and segment characteristics.

Dataset Insights

Analysis typically reveals 5 customer segments: balanced spenders, high-income low-spenders, conservative shoppers, target high-value customers, and young high-spenders. These insights support personalized marketing, inventory planning, and customer retention strategies.

Requirements

Python 3.6+
Libraries: pandas, numpy, matplotlib, seaborn, scikit-learn
Environment: Google Colab (uses google.colab.files for upload) or local Jupyter with equivalent file handling.

No additional installations needed in Colab as libraries are pre-installed.

Usage

Open in Google Colab:

• Go to Google Colab.
• Upload the notebook: File > Upload notebook > Select "Customer-Segmentation.ipynb".

Prepare Data:
Ensure you have a CSV file with customer data. Expected columns include:

• CustomerID: Unique customer identifier.
• Gender: Customer gender (Male/Female).
• Age: Customer age.
• Annual Income: Annual income in thousands.
• Spending Score: Score assigned by mall (1-100 based on behavior).

Example data source: Kaggle Mall Customer Segmentation dataset.

Run the Notebook:

• Execute cells sequentially (Shift+Enter).
• In the upload cell, provide your CSV when prompted.
• The notebook performs EDA, finds optimal clusters, applies K-Means, and displays segmentation plots.

Expected Outputs:

• Console prints: Dataset shape, feature statistics, optimal k value, silhouette score.
• Plots:
  • Elbow method for optimal k.
  • Silhouette analysis.
  • Cluster scatter plot (Income vs Spending Score).
  • Segment distribution visualizations.