Aurelie Giraud
Data Scientist & Data Product Manager
LinkedIn | Blog
Hi 👋 I'm Lily,
Data Scientist, aspiring storyteller, and writer - on a mission to democratize Analytics. It all started during my Master's degree when I discovered a passion for Statistics and Experimental Design. Shortly after - as a project manager, I learned how businesses work and how to communicate with stakeholders from disparate backgrounds. Ultimately I realized that I get the most joy from making a business impact based on data, analytics, and automation. From there, I decided to drop everything and start over to become a Data Scientist. Today, when I'm not coding, you can find me writing blogs on how to apply Data Science to solve business challenges and self-growth.
Customer Segmentation based on credit card data
✅ Business Case
Create a segmentation of customers based on their credit card utilization for improving the marketing strategy for a bank.
📉 Dataset
The data can be found in Kaggle: https://www.kaggle.com/arjunbhasin2013/ccdata
📒 Repository
The repository is available on GitHub. This project includes the following steps:
- Data Wrangling
- Data Cleaning
- Handling outliers
- Dealing with missing values
- Check correlations and distributions
- Log-transformation
- Segmentation
- Preprocessing: normalization, tuning,…
- Select most important features to highlight the key criteria to segment the customers.
- Clustering with various ML algorithms: K-Means, SpectralClustering, AgglomerativeClustering, DBSCAN and K-Medoid.
- Visualization of cluster’s characterictics to understand the typology of customers.
- Cluster Analysis based on distance between the centroids and statistical tests (Mann-Whitney) to check if all clusters are significant.