Bayesian Differential Privacy for Linear Dynamical Systems

Differential privacy is a measure based on the difficulty of discriminating between similar input data. In differential analysis, data usually implies that their distance does not exceed predetermined threshold. It, consequently, take into account distinguishing sets are far apart, which often contain highly private information. This problem has been pointed out in research for static data, and Bayesian proposed, provides protection level even outlier by utilizing prior distribution this study, we introduce to dynamical systems, provide guarantees distant pairs reveal its fundamental property. For example, design mechanism satisfies desired protection, characterizes trade-off information utility.