/PRODIGY_ML_02

Primary LanguageJupyter Notebook

PRODIGY_ML_02

ML Internship at Prodigy InfoTech

Task - 02

K-means clustering algorithm to group customers

This repository contains code on K-means clustering algorithm to group customers of a retail store based on their purchase history.

Description:

In the dynamic landscape of retail, understanding customer behavior is paramount for business success. Leveraging the power of machine learning, this project focuses on employing the K-means clustering algorithm to categorize customers of a retail store based on their purchase history. The aim is to uncover meaningful patterns in customer data, allowing the retail store to tailor its strategies for different customer segments.

Algorithm:

Step 1: Import Necessary Libraries

  • Import required libraries such as pandas, numpy, matplotlib.pyplot, seaborn, and KMeans from sklearn.

Step 2: Load and Explore the Data

  • Read the 'Mall_Customers.csv' dataset into a DataFrame (df).
  • Display the first few rows of the dataset using df.head().
  • Display the shape of the dataset using df.shape.
  • Display information about the dataset using df.info().

Step 3: Prepare Data for Clustering

  • Select relevant columns (features) for clustering, typically representing customer behavior. In this case, columns 3 and 4 are selected and stored in the variable 'X'.

Step 4: Elbow Method for Optimal Cluster Selection

  • Initialize an empty list 'wcss' to store Within-Cluster-Sum-of-Squares.
  • Iterate through a range of cluster numbers (1 to 10) using a for loop.
  • For each cluster number, fit a KMeans model and calculate the WCSS, then append it to the 'wcss' list.
  • Plot an elbow chart to visualize the WCSS values.
  • Choose the number of clusters where the decrease in WCSS starts to slow down (the "elbow" point). In this case, it seems to be 5.

Step 5: Perform K-means Clustering

  • Initialize a KMeans model with the chosen number of clusters (5).
  • Fit the model on the data and obtain cluster labels for each data point.

Step 6: Visualize the Clusters

  • Plot a scatter plot for the customer data points, color-coded by their assigned clusters (Cust 1 to Cust 5).
  • Plot the centroids of the clusters in magenta.
  • Add a legend to identify each cluster.

Step 7: Display the Plot

  • Display the final plot.

Usage:

  1. Dataset from Kaggle Mall Customer Segmentation Data.
  2. Used Jupiter Notebook for Python Coding.

Techniques:

The following techniques are implemented in this project:

  • K-means clustering algorithm