Multiparty Data Publishing via Blockchain and Differential Privacy
Algorithm 1
LDA-DP algorithm.
Input: Data sets , privacy parameters ,
Output: Projection direction vector , projection data
(1)
for to 2 do
(2)
Set , which generates a dimension noise vector ; each entry is sampled from
(3)
Computes
(4)
end for
(5)
return
(6)
Set , which generates a random matrix . Let be a symmetric matrix with the upper triangle (including the diagonal) entries are sampled from and make the symmetrical position entries in the lower triangle matrix equal to the upper triangle.