Answer: (ii) Dynamic programming reduces the time complexity by avoiding redundant computations.

Explanation:

• Brute-force / naive recursion (e.g., recursive Fibonacci, naive matrix chain) recomputes same subproblems exponentially many times.
• DP stores results of overlapping subproblems in a table (memoization or tabulation) so each distinct subproblem is solved once → transforms exponential to polynomial (e.g., Fibonacci $O(2^n) \to O(n)$, LCS $O(2^{m+n}) \to O(mn)$).

Why not other options:

• (i) Both brute-force and DP can find global optimum if they explore whole space; DP does not uniquely guarantee it—correctness comes from optimal substructure.
• (iii) Both divide-and-conquer and DP divide into subproblems; DP's distinguishing advantage is memoization of overlapping subproblems, not mere division.
• (iv) DP is often less intuitive due to recurrence and table ordering.

Correct choice: (ii)