Unsupervised ML Customer Segmentation

Problem Description

In this project, your task is to identify major customer segments on a transnational data set which contains all the transactions occurring between 01/12/2010 and 09/12/2011 for a UK-based and registered non-store online retail.The company mainly sells unique all-occasion gifts. Many customers of the company are wholesalers.

Step used for Analysis

• Data Cleaning
  
• Exploratory Data Analysis (EDA)
  
• RFM
  
• Clustering
  
• Cluster Profiling
  
• Conclusion
  

Recency-Frequency-Monetary (RFM) model to determine customer value:

The RFM model is quite useful model in retail customer segmentation where only the data of customer transaction is available. RFM stands for the three dimensions:

Recency – How recently did the customer purchase? - Number of days since last purchase
Frequency – How often do they purchase? - The total number of purchases
Monetary Value – How much do they spend? - The total money customer spent
A combination of these three attributes can be defined to assign a quantitative value to customers. e.g. A customer who recently bought high value products and transacts regularly is a high value customer.

Model used for analysis

• K-Means Clustering
  
  • Silhouette Analysis
  • Elbow Method
• DBSCAN Clustering
• Hierarchical Clustering
  
  • Agglomerative Clustering (Bottom-up approach using tree like diagram called Dendogram)

Conclusion

• RFM(Recency, Frequency and Monetary) dataframe ease our problem to solve in a particular order, it makes easy to recommend and display new launched products to few customers.
• And total number clusters for each of the used models came 2.

Team Members

Sonica Sinha
Mohd Taufique

Acknowledgements

© Copyright