🌟 Customer Segmentation Project 📊

Welcome to the Customer Segmentation Project! This project uses KMeans clustering to identify customer segments and provides an interactive Streamlit dashboard for data visualization. Explore the code, analyze clusters, and gain insights into customer behaviors. 🛍️

---

📁 Project Structure

customer-segmentation-project/
├── data/                            # Directory for data files
│   ├── customers.csv                # Raw dataset
│   ├── customers_preprocessed.csv   # Preprocessed dataset
├── src/
│   ├── data_preprocessing.py        # Data cleaning, encoding, and scaling script
│   ├── clustering.py                # KMeans clustering and evaluation script
│   ├── utils.py                     # Utility functions for data handling
│   ├── visualization.py             # PCA and cluster visualization functions
├── app/
│   ├── dashboard.py                 # Streamlit dashboard for interactive visualization
├── case study/
│   ├── R&D(ML)_DeeptiGupta.pdf      # Case study documentation
├── images/                          # Images for Readme documentation
│   ├── preprocessing.png            # Visualization of preprocessing steps
│   ├── clustering.png               # Example of clustering output
│   ├── dashboard.png                # Streamlit dashboard view
├── requirements.txt                 # Python dependencies
└── README.md                        # Project documentation

---

🔧 Setup and Installation

1. Clone the Repository:

   git clone https://github.com/Deeptig9138/customer-segmentation-project.git
   cd customer-segmentation-project
   

2. Install Dependencies:

   pip install -r requirements.txt
   

3. Run the Dashboard:

   streamlit run app/dashboard.py
   

---

📝 Project Workflow

1. Data Preprocessing 🧹

The data_preprocessing.py script handles missing values, encodes categorical variables, and scales numerical data using StandardScaler. It saves the cleaned data for clustering.

Sample Code:

import pandas as pd
from sklearn.preprocessing import StandardScaler, LabelEncoder

def load_data(file_path):
    return pd.read_csv(file_path, sep='\t')

Output: A preprocessed CSV file saved in the data/ folder.

2. Clustering 🤖

The clustering.py script applies KMeans clustering and evaluates the performance using the Silhouette Score.

Sample Code:

from sklearn.cluster import KMeans
from sklearn.metrics import silhouette_score

def perform_kmeans_clustering(df, n_clusters):
    kmeans = KMeans(n_clusters=n_clusters, random_state=42)
    labels = kmeans.fit_predict(df)
    return labels, kmeans

3. Interactive Dashboard 📈

Use Streamlit in dashboard.py to visualize clusters interactively. The dashboard includes PCA plots for dimensionality reduction and cluster interpretation.

Dashboard Features:

• View sample data.
• Adjust the number of clusters dynamically.
• Visualize clusters in a 2D PCA plot.

Run the dashboard using the command below:

streamlit run app/dashboard.py

📊 Visualizing Clusters

The visualization.py script leverages PCA for dimensionality reduction and uses Seaborn for beautiful scatter plots.

Sample Visualization Code:

import matplotlib.pyplot as plt
import seaborn as sns

def plot_clusters(df, labels):
    sns.scatterplot(x=reduced_data[:, 0], y=reduced_data[:, 1], hue=labels, palette="viridis")
    plt.show()

⚙️ Utility Functions

The utils.py script contains helper functions for saving and loading data:

• save_data: Save data to a CSV file.
• load_processed_data: Load preprocessed data.

---

📂 Case Study 📖

A detailed Case Study is provided in the case study folder.

• Explains the problem statement
• Answer the given questions in the task 1.

---

🎯 Key Results

• Silhouette Score: Measure of cluster quality.
• PCA Visualization: Easy interpretation of clusters in 2D space.
• Interactive Dashboard: Real-time cluster exploration.

---

🛠️ Technologies Used

• Python 🐍
• Pandas & NumPy
• Scikit-learn
• Streamlit
• Matplotlib & Seaborn

---

🤝 Contributions

Feel free to submit pull requests or report issues! Contributions are always welcome. 🙌

💡 Thank you for exploring the Customer Segmentation Project! 💡