Customer Personality Analysis & Segmentation

Customer segmentation analysis using clustering algorithms to identify distinct customer groups for targeted marketing strategies.

📁 Dataset

Source: Customer Personality Analysis Dataset
Context: 2,240 customer records with 29 behavioral features
Attributes Include:

• Demographics (Age, Education, Marital Status, Income)
• Purchasing Behavior (Wines, Fruits, Meat, Fish, Sweets)
• Campaign Responses (AcceptedCmp1-5)
• Engagement Metrics (Recency, Web Visits, Store Purchases)

🛠️ Key Steps(Full report in Report.pdf)

1. Data Preprocessing

• Handled missing values
• Feature engineering
• Encoded categorical features
• Removed outliers

2. Feature Standardization

• Scaled features using StandardScaler
• Ensured equal weight for income (high magnitude) vs binary features

3. Clustering & PCA

Algorithms Tested:

• K-Means
• BIRCH
• MiniBatchKMeans
• Agglomerative Clustering
• Mean Shift Dimensionality Reduction:
• PCA achieved 90% variance retention with 7 components
• Silhouette Score Comparison

4. Cluster Visualization

📊 Key Findings

3 Distinct Customer Segments Identified:

1. Cluster 0 - Budget Families

   • Lowest income (price-sensitive)
   • Middle-aged with children
   • Prefers deals on Gold/Fish/Sweets

2. Cluster 1 - Affluent Professionals

   • Highest income (top 15%)
   • Young, educated, no kids
   • Wine/Meat enthusiasts, catalog shoppers

3. Cluster 2 - Conservative Seniors

   • Moderate income
   • Oldest demographic (many widowed)
   • Low campaign engagement