A Note on Parameters of Random Substitutions by .GAMMA.-Diagonal Matrices
Ju-Sung Kang · IEICE Transactions on Fundamentals of Electronics Communications and Computer Sciences · 2010
Random substitutions are very useful and practical method for privacy-preserving schemes. In this paper we obtain the exact relationship between the estimation errors and three parameters used in the random substitutions, namely the privacy assurance metric γ, the total number n of data records, and the size N of transition matrix. We also demonstrate some simulations concerning the theoretical result.