1 动态规划(变相最长公共子序列)
【本题特点】:dp
数组不仅能存储子序列的长度,还能存储子序列本身,确实可以很灵活
dp[i][j] = dp[i - 1][j - 1] + s1[i - 1];
dp[i][j] = max(dp[i - 1][j], dp[i][j - 1]);
class Solution {
public:
int minimumDeleteSum(string s1, string s2) {
int asize = s1.size();
int bsize = s2.size();
vector<vector<int>> dp(asize + 1, vector<int>(bsize + 1, 0));
for (int i = 1; i <= asize; i++)
for (int j = 1; j <= bsize; j++) {
if (s1[i - 1] == s2[j - 1])
dp[i][j] = dp[i - 1][j - 1] + s1[i - 1];
else
dp[i][j] = max(dp[i - 1][j], dp[i][j - 1]);
}
return accumulate(s1.begin(),s1.end(),0) + accumulate(s2.begin(),s2.end(),0) - 2*dp[asize][bsize];
}
};