Robin Steinkühler

  School of Business and Economics | Bachelor Business Analytics

"Curating Privacy Concerns: Optimizing Synthetic Data Generation for Machine Learning in Healthcare"

Robin's elevator pitch
"Artificial Intelligence (AI) is becoming increasingly promising in healthcare. However, ensuring patient privacy remains a major concern that limits the use of AI in this field. This thesis proposes synthetic data, which not only mimics the structure and statistical properties of patient data without the risk of revealing sensitive information but can also be generated in large volumes, as a viable solution to mitigate privacy concerns. The results underscored the high utility of synthetic data, with a 10% improvement in AUC ROC scores of a decision tree classifier trained on synthetic data compared to one trained on real patient data."