The greedy approach is optimal for the Fractional Knapsack problem, as it can always make the locally optimal choice to maximize the total value. For problems like the 0/1 Knapsack, the greedy approach does not guarantee an optimal solution. The Longest Increasing Subsequence and Matrix Chain Multiplication problems require dynamic programming for optimal solutions.