A collaborative mini-research project analyzing Wasserstein GANs (WGANs) through extensive literature review and experimental evaluation. Explores training stability, loss behavior, gradient penalties, and convergence characteristics, proposing insights to improve generative model robustness.

A Python pipeline for segmenting financial assets using unsupervised learning models like K-Means and GMM. This project evaluates clustering configurations not only on performance (Silhouette Score) but also on their statistical stability across train, test, and validation sets using the Wasserstein distance.

Event-driven parameter anchoring experiments for post-convergence model stability under uncertain feedback.

Adult Income Drift Lab conducts a comprehensive model stability analysis under demographic covariate shift, combining statistical drift detection with performance and calibration evaluation on real-world census data.

🌊 Explore Wasserstein GANs through in-depth research and practical insights to enhance stability and performance in generative modeling.