Search code examples
algorithmtime-complexitycomplexity-theory

Does 2 ^ O(log log n) = O(log n)?


Does 2 ^ O(log log n) = O(log n)? Could you explain how to test this relationship?

I have tried substituting O(log(log n)) by C1 log(log n) and log n by C2 log n to find the relationship between them. When I graph the functions it seems that the statement is true, but I get stuck in the mathematical proof at one point and am not sure how to proceed.


Solution

  • No. Assuming a base-2 log, then

    2log log N = log N,

    but 210 log log N is also in 2O(log log N), and

    210 log log N = (2log log N)10 = (log N)10

    ...and that is obviously not in O(log N)