Back

Improving Privacy And Utility In Aggregate Data: A Hybrid Approach Using Pooldiv

Zhang, I. M.; Chen, K.; Percy, W.

2024-05-06 health informatics
10.1101/2024.05.05.24306903 medRxiv
Show 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. Through extensive simulations and real data analysis, we assess the performance of synthetic datasets generated via traditional DP methods, data pooling, and our proposed PoolDiv method, demonstrating the advantages of our hybrid approach in maintaining data utility while ensuring privacy.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.