What is time complexity of
$$T(n)=4 \cdot T \left(\frac{\sqrt{n}}{3} \right) + \log^2 n$$
I cannot solve it by recursion tree or any other method.
we have the base T(1) = 1
I may think about sandwich theorem but dont know how to use.
What is time complexity of
$$T(n)=4 \cdot T \left(\frac{\sqrt{n}}{3} \right) + \log^2 n$$
I cannot solve it by recursion tree or any other method.
we have the base T(1) = 1
I may think about sandwich theorem but dont know how to use.
On
The following is an upper bound, which probably is good enough for any analysis.
We have $$T(n)\leq4 \cdot T \left(\frac{n}{3} \right) + n$$ Using Master Theorem we get $O(n^{\log_34})$.
Let $b$ be the base of the logarithm, and let $n = b^m$. Then
$$ T(b^m) = 4 \cdot T\left( b^{m/2 - \log 3} \right) + \log^2 n $$
which suggests setting
$$ S(m) = T(b^m) \qquad \qquad T(n) = S(\log n)$$
and then first solving the recurrence
$$ S(m) = 4 \cdot S\left(\frac{m}{2} - \log 3 \right) + m^2 $$
where I've assumed by $\log^2 n$ you mean $(\log n)^2$ rather than $\log \log n$.