Multidomain Fusion Data Privacy Security Framework

Algorithm 3

Group reconstruction.

Input: the constraint inference frequency, L[i]: the original histogram frequency;

Output:the grouping reconstruction result;

1 define an array DV[i] to store ;

2 define the SS, SSEr, SSEl store the error between the grouped histogram and the original histogram;

3 define avgR, avgL, avg store the average of the group histogram;

4 for i∈[0,k]  // Lines 4-6 calculate the absolute value of the difference between adjacent numbers;

5 Dv[i]=abs(]-);

6 end for

7 ;  // Calculate the average frequency when all buckets are combined into a group;

8 calculate the SSE after merging a group;

9 for each i in Dv[i]   // Line 9-28 Group reconstruction based on SSE

10 if (sum(DV[i]));  // group terminal condition;

11  break;

12 else

13  mid = getMaxId();  // get the max DV location value

14  DV[mid]=-1;

15  left getLeft(mid);   // get the left boundary according to the max DV

16  right getRighr(mid); // get the right boundary according to the max DV

17  flag[mid]=true;

18  for i∈[left, mid]

19   avgR = ;

20   SSEr = (L[i]-avgR)²;

21  end for

22  for i∈[mid+1,right];

23   avgL = ;

24   SSEl = (L[i]-avgL)²;

25  end for

26  SSE[i] SSEr+SSEl;

27  record the result of the current grouping reconstruction;

28 end for

29 sort(SSE[i]) and select the minimum SSE;

30 return the grouping reconstruction of the minimum SSE