Shopping Data Analysis & Customer Segmentation

This project performs analysis on a shopping dataset to derive insights about sales trends, customer behavior, and product performance. The key techniques used include data preprocessing, clustering, market basket analysis, and visualization.

Features:

1. Data Preprocessing:
   • Handling missing values
   • Calculating total sales for each transaction
2. Sales Analysis:
   • Top products by sales
   • Monthly sales trend
3. Customer Segmentation:
   • KMeans clustering to segment customers based on total sales and quantity purchased
4. Market Basket Analysis:
   • Using Apriori algorithm to find frequent itemsets and association rules

Libraries Used:

• pandas for data manipulation
• numpy for numerical operations
• matplotlib and seaborn for data visualization
• sklearn for machine learning models (KMeans, StandardScaler)
• mlxtend for Market Basket Analysis (Apriori algorithm and Association Rules)

Installation:

To run this project locally, you'll need Python 3.x and the following libraries:

pip install pandas numpy matplotlib seaborn scikit-learn mlxtend

Dataset:

The dataset (shopping_data.csv) should be placed in the project directory. This dataset contains shopping transactions with columns like InvoiceDate, Description, Quantity, UnitPrice, and CustomerID.

Code Walkthrough:

1. Data Preprocessing:

   • The data is loaded and missing values are removed.
   • The TotalSales for each transaction is calculated as the product of Quantity and UnitPrice.

2. Sales Analysis:

   • The top 10 products by total sales are displayed.
   • Monthly sales trends are visualized with a line plot.

3. Customer Segmentation:

   • The data is scaled using StandardScaler.
   • KMeans clustering is applied to segment customers based on total sales and quantity purchased.
   • A scatter plot is created to visualize customer clusters.

4. Market Basket Analysis:

   • A binary matrix is created to represent the items purchased in each transaction.
   • Apriori algorithm is used to find frequent itemsets, and association rules are derived.
   • A scatter plot visualizes the association rules with metrics like support, confidence, and lift.

Visualizations:

• Monthly Sales Trend: Line plot showing the total sales over time.
• Top 10 Products by Sales: Bar chart showing the products that contribute the most to sales.
• Customer Segmentation: Scatter plot displaying customer clusters.
• Correlation Heatmap: Heatmap showing correlations between quantity, unit price, and total sales.
• Association Rules: Scatter plot visualizing the association rules with metrics such as support, confidence, and lift.

Example Outputs:

• Top 10 Products by Sales:

  Description        TotalSales
  Product A          5000
  Product B          4000
  ...
  

• Customer Segments:

  Cluster  TotalSales  Quantity
  0        15000       200
  1        12000       150
  ...
  

• Association Rules:

  Rule 1: {Product A} -> {Product B} (Lift: 1.5)
  Rule 2: {Product C} -> {Product D} (Lift: 2.0)
  ...
  

Usage:

Run the script after placing the shopping_data.csv file in the directory. The output will include statistical summaries, visualizations, and insights about customer behavior and product sales.

License:

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