A1434
Title: Mitigating bias in analyzing privacy-preserved time-to-event data
Authors: Yi Xiong - University at Buffalo (United States) [presenting]
Abstract: Sharing time-to-event data is beneficial for enabling collaborative research efforts, facilitating the design of effective interventions, and advancing patient care. However, sharing the exact survival curves poses concerns over privacy. Although there are several popular privacy-protecting solutions (e.g., binning, differential privacy) that offer strong protection on the data, the "sanitized" data usually has low utility and can result in misleading statistical inference. The aim is to investigate the distortion in bias and variance in regression analysis of sanitized survival data under popular privacy-protecting solutions, and to provide a strategy to mitigate the bias in estimators with sanitized survival data.
