End to end ML project predicting cubic zirconia prices from physical attributes (cut, color, clarity, dimensions). Walks the full pipeline from EDA through feature engineering to model evaluation.
What I did
Achieved 97.8% accuracy on the held-out test set using regression with engineered features.
Full EDA covering distributions, correlations, and outlier handling to inform modeling choices.
Feature engineering for non-linear interactions between physical attributes and price.
Model comparison across linear regression, decision trees, and ensemble methods to justify the final choice.
