Two data sets: offers details and customer transactions associated with offers.
This project utilizes Sklearn to demonstrate various clustering techniques including K-means, DBSCAN, Spectral Clustering, Affinity Propagation, and Agglomerative Clustering. Relevant methods for selecting number of clusters in K-means, namely the Elbow method and the Silhouette Coefficient were also tested.