Explainable AI for Medical Data

Context:

Developed as part of research and applied academic work on

real medical datasets.

Challenge:

Medical data requires high accuracy and clear interpretability

—black-box models are not acceptable.

My Approach:

Designed an end-to-end ML workflow including preprocessing,

feature engineering, model training, and explainability using

SHAP.

Outcome:

Transparent predictions suitable for sensitive domains

Clear feature-level insights for decision-makers