ABSTRACT The increasing need to protect individual privacy in data releases has led to significant advancements in privacy‐preserving technologies. Differential Privacy (DP) offers robust privacy guarantees but often at the expense of data utility. On the other hand, data pooling, while improving utility, lacks formal privacy assurances. Our study introduces a novel hybrid method, termed PoolDiv, which combines differential privacy with data pooling to enhance both privacy guarantees and data utility. PoolDiv achieves this balance by adopting relaxed ()‐differential privacy parameters and optimizing pooling sizes (e.g., ), thereby enhancing privacy without significantly compromising utility. Using datasets such as the Lymphoma microRNA data and simulated synthetic data, we evaluate PoolDiv's performance through metrics such as R‐squared (), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE). PoolDiv's optimized pooling strategy preserves crucial data structures, enabling more accurate and insightful analyses compared to traditional DP. For example, in healthcare research, PoolDiv can facilitate the secure release of patient data for disease prediction while preserving complex relationships between biomarkers. Similarly, in financial systems, PoolDiv enables fraud detection without compromising the privacy of individual transactions. Although computational overhead and over‐smoothing with larger pooling sizes pose challenges, our findings demonstrate PoolDiv's effectiveness in preserving underlying data structures and providing actionable insights. This approach addresses critical gaps in existing privacy‐preserving technologies, paving the way for secure and practical data‐sharing practices across sensitive domains.
Kofie et al. (Sun,) studied this question.