动态规划总结（Dynamic programming summary）.docVIP

动态规划总结（Dynamic programming summary） Although all are used to the previous results But the way of dealing is very different. Greed is the best choice for each one, producing an optimal result The dynamic programming is to calculate the optimal results produced by all the choices at each step. Finally, the optimal results are obtained ----------------------------------------------------------------- Greedy algorithm is a strategy, thinking... It does not have a fixed pattern For example, the simplest knapsack problem There are probably many ways to do it with greed Of the highest value, value, or weight And you cant make sure that the greedy policy you choose is absolutely optimal for all situations The idea of dynamic programming is divide and conquer + solve sink Break a complex problem into small pieces The optimal solution is obtained for every small problem And get better answers from these optimal solutions Typical examples of number tower problems Draw a picture and youll see it ----------------------------------------------------------------- The difference between dynamic programming and greedy algorithms Dynamic programming and greedy algorithm are recursive algorithms The local optimum solution is used to derive the global optimum Difference： Greedy algorithm: 1. the greedy algorithm, make every step of the greedy decision can be changed, because the greedy strategy is the optimal one step optimal solution is the next step of the solution, and the optimal solution before a no reservation. 2. from the introduction of (1), we can know that the correct condition of greedy method is that the optimal solution of each step must contain the best solution of the previous step. Dynamic programming algorithm: 1., the global optimal solution must contain a local optimal solution, but it does not necessarily contain the previous local optimal solution. Therefore, all the prior optimal solutions need to be recorded 2., the key of dynamic programming is the state transit