Agglomerative Clustering

Dataset Description The dataset agglomerative_clustering_dataset.csv contains 5,000 customer observations generated to simulate real-world retail customer behavior for the application of agglomerative hierarchical clustering. Each observation represents a unique customer and includes demographic, financial, and behavioral attributes that are commonly used in customer segmentation studies. The dataset is designed for educational and research purposes, allowing students to explore hierarchical clustering techniques, evaluate cluster structures, and interpret customer groups using both statistical analysis and machine learning methods. Its multidimensional nature makes it suitable for demonstrating the complete agglomerative clustering workflow, including data preprocessing, feature scaling, dendrogram construction, cluster formation, and cluster interpretation. The dataset consists of the following variables: Variable Description CustomerID Unique identifier assigned to each customer. Annual Income Annual customer income, expressed in thousands of U.S. dollars (USD). SpendingScore A score ranging from 1 to 100 that measures the customer's purchasing behavior, where higher values indicate greater spending activity. Age Customer age in years. Online Purchases Total number of online purchases made during the observation period. Store Visits Number of visits made to physical retail stores. LoyaltyScore A score between 0 and 100 representing customer loyalty based on purchase frequency and engagement. Credit Score Customer credit rating, ranging approximately from 450 to 850, reflecting overall financial reliability. This dataset is particularly suitable for hierarchical clustering because it contains variables measured on different numerical scales, highlighting the importance of feature standardization before clustering. It enables students to investigate customer similarities, generate dendrograms, identify natural customer segments, compare hierarchical clustering with partition-based methods such as K-Means, and gain practical experience in applying unsupervised learning techniques to realistic business data.