←

Customer Clustering

unsupervised machine learning // python // scikit-learn // data visualization

A customer segmentation project using K-Means, feature engineering, and PCA visualizations to identify behavioral patterns in the Instacart dataset.

This project explores how unsupervised learning can be used to segment customers based on shopping behavior. Using the Instacart dataset, I built a full K-Means clustering pipeline with feature engineering, scaling, and visualization to evaluate cluster separation and interpret emerging customer groups.

development

Built a full preprocessing pipeline including feature engineering, scaling, and missing-data handling
Applied K-Means clustering with Elbow Method and Silhouette Analysis to determine optimal cluster count
Used PCA, heatmaps, and parallel coordinate plots to validate cluster separation and interpret behavior patterns
Identified meaningful customer profiles (e.g., frequent vs. bulk shoppers) based on cluster characteristics
Assessed ethical considerations around segmentation and model-driven decision-making

technologies

Python · scikit-learn · Pandas · NumPy · Matplotlib · Seaborn · PCA · K-Means Clustering · Data Preprocessing · Feature Engineering · Jupyter Notebook

View repo →