Machine Learning
Customer Segmentation and Strategic Recommendation
K-Means segmentation for investment propensity strategy
Overview
A customer analytics project that uses multivariate EDA and K-Means clustering to identify customer segments and recommend targeted investment strategies.
Problem
Financial product strategy needs actionable customer groups rather than broad averages across a mixed population.
Dataset
A 1,000+ record customer dataset with behavioral and demographic variables used for segmentation and reporting.
Approach
Performed multivariate EDA, engineered clustering inputs, trained Scikit-learn K-Means models, and validated segment separation with Silhouette scores.
Results
Identified five customer segments and proposed strategies that increased investment propensity by 10% compared with baseline.
Lessons Learned
Segmentation is strongest when the model output is translated into concrete recommendations that a business team can act on.
Model / Pipeline
The implementation combines Python, Scikit-learn, Pandas, NumPy, Matplotlib, Seaborn with a repeatable workflow for data preparation, evaluation, and communication.