Welcome to the Olist Customer Segmentation and Client Clustering project repository! The aim is to understand different client profiles from the Olist client dataset. The project explores different unsupervised learning approaches.

contains a detailed data exploration and cleaning process.

various clustering approaches to achieve customer segmentation and client profiling:

• K-Means Clustering: Utilizing RFM (Recency, Frequency, Monetary values) and additional features like localization, satisfaction, and payment installments for segmentation.
• PCA-Enhanced Clustering: Incorporating Principal Component Analysis (PCA) to optimize segmentation results.
• Alternative Clustering Algorithms: Exploring Agglomerative Clustering and DBSCAN for a deeper understanding of client profiles.

We recommend how often to update the segmentation using the Adjusted Rand Index (ARI) to ensure its relevance. The study encompasses:

For further insights into our project, including detailed explanations in French, please refer to the presentation slides provided in PDF format.