HDU 1693 新的恐惧,插头Dp

Problem Description
Most of us know that in the game called DotA(Defense of the Ancient), Pudge is a strong hero in the first period of the game. When the game goes to end however, Pudge is not a strong hero any more.
So Pudge’s teammates give him a new assignment—Eat the Trees!

The trees are in a rectangle N * M cells in size and each of the cells either has exactly one tree or has nothing at all. And what Pudge needs to do is to eat all trees that are in the cells.
There are several rules Pudge must follow:
I. Pudge must eat the trees by choosing a circuit and he then will eat all trees that are in the chosen circuit.
II. The cell that does not contain a tree is unreachable, e.g. each of the cells that is through the circuit which Pudge chooses must contain a tree and when the circuit is chosen, the trees which are in the cells on the circuit will disappear.
III. Pudge may choose one or more circuits to eat the trees.

Now Pudge has a question, how many ways are there to eat the trees?
At the picture below three samples are given for N = 6 and M = 3(gray square means no trees in the cell, and the bold black line means the chosen circuit(s))


 

Input
The input consists of several test cases. The first line of the input is the number of the cases. There are no more than 10 cases.
For each case, the first line contains the integer numbers N and M, 1<=N, M<=11. Each of the next N lines contains M numbers (either 0 or 1) separated by a space. Number 0 means a cell which has no trees and number 1 means a cell that has exactly one tree.
 

Output
For each case, you should print the desired number of ways in one line. It is guaranteed, that it does not exceed 2 63 – 1. Use the format in the sample.
 

Sample Input
  
  
2 6 3 1 1 1 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 2 4 1 1 1 1 1 1 1 1
 

Sample Output
  
  
Case 1: There are 3 ways to eat the trees. Case 2: There are 2 ways to eat the trees.
 

题目大意就是求完全覆盖的合法方案数。

插头Dp。感受到了博弈论时候的恐惧QAQ。

那么我的f[i][j][S]表示处理完(i,j)这个格子后插头的集合QWQ。

那么f[i][0][S<<1]=f[i-1][m][S];因为枚举处理完上一行的状态来更新这一行的状态。

f[i][j][k]=f[i][j-1][k^(1<<(j-1))^(1<<j)]考虑每个方块有一个或两个的都能考虑进去(除去单个竖直插头)

f[i][j][k]=f[i][j-1][k]当这个方块只有一个竖直插头的时候可以由上一个方块转移。

上面讨论的都是非障碍,下面讨论障碍。

如果当前障碍(i,j)上没有插头则继承上一个即f[i][j][S]=f[i][j-1][S];

然后答案就是f[n][m][0]即操作完(n,m)后剩余插头状态为0(因为轮廓线下移。)

code:

#define ll long long


#include <stdio.h>
#include <cstring>
#include <iostream>


template<typename _t>
inline _t read(){
    _t x=0,f=1;
    char ch=getchar();
    for(;!isdigit(ch);ch=getchar())if(ch=='-')f=-f;
    for(;isdigit(ch);ch=getchar())x=x*10+(ch^48);
    return x*f;
}

int n,m,a[13][13],Tcase,num;
ll f[13][13][1<<13];

void Dp(){
    memset(f,0,sizeof f);
    f[0][m][0]=1;
    int full = 1<<(m+1);
    for(int i=1;i<=n;i++){
        for(int j=0;j<full>>1;j++)
            f[i][0][j<<1]=f[i-1][m][j];
        for(int j=1;j<=m;j++)
            for(int k=0;k<full;k++){
                int x = 1<<(j-1);
                int y = 1<<(j);
                if(a[i][j]){
                    f[i][j][k]+=f[i][j-1][k^x^y];
                    if((k&x)&&(k&y))continue;
                    if(!(k&x)&&!(k&y))continue;
                    f[i][j][k]+=f[i][j-1][k];
                }
                else{
                    if(!(k&x)&&!(k&y))f[i][j][k]+=f[i][j-1][k];
                    else f[i][j][k]=0;
                }
            }
    }
}

int main(){
    Tcase=read<int>();
    while(Tcase--){
        n=read<int>(),m=read<int>();
        for(int i=1;i<=n;i++)
            for(int j=1;j<=m;j++)
                a[i][j]=read<int>();
        Dp();
        printf("Case %d: There are %lld ways to eat the trees.\n",++num,f[n][m][0]);
    }
}





深度学习是机器学习的一个子领域,它基于人工神经网络的研究,特别是利用多层次的神经网络来进行学习和模式识别。深度学习模型能够学习数据的高层次特征,这些特征对于图像和语音识别、自然语言处理、医学图像分析等应用至关重要。以下是深度学习的一些关键概念和组成部分: 1. **神经网络(Neural Networks)**:深度学习的基础是人工神经网络,它是由多个层组成的网络结构,包括输入层、隐藏层和输出层。每个层由多个神经元组成,神经元之间通过权重连接。 2. **前馈神经网络(Feedforward Neural Networks)**:这是最常见的神经网络类型,信息从输入层流向隐藏层,最终到达输出层。 3. **卷积神经网络(Convolutional Neural Networks, CNNs)**:这种网络特别适合处理具有网格结构的数据,如图像。它们使用卷积层来提取图像的特征。 4. **循环神经网络(Recurrent Neural Networks, RNNs)**:这种网络能够处理序列数据,如时间序列或自然语言,因为它们具有记忆功能,能够捕捉数据中的时间依赖性。 5. **长短期记忆网络(Long Short-Term Memory, LSTM)**:LSTM 是一种特殊的 RNN,它能够学习长期依赖关系,非常适合复杂的序列预测任务。 6. **生成对抗网络(Generative Adversarial Networks, GANs)**:由两个网络组成,一个生成器和一个判别器,它们相互竞争,生成器生成数据,判别器评估数据的真实性。 7. **深度学习框架**:如 TensorFlow、Keras、PyTorch 等,这些框架提供了构建、训练和部署深度学习模型的工具和库。 8. **激活函数(Activation Functions)**:如 ReLU、Sigmoid、Tanh 等,它们在神经网络中用于添加非线性,使得网络能够学习复杂的函数。 9. **损失函数(Loss Functions)**:用于评估模型的预测与真实值之间的差异,常见的损失函数包括均方误差(MSE)、交叉熵(Cross-Entropy)等。 10. **优化算法(Optimization Algorithms)**:如梯度下降(Gradient Descent)、随机梯度下降(SGD)、Adam 等,用于更网络权重,以最小化损失函数。 11. **正则化(Regularization)**:技术如 Dropout、L1/L2 正则化等,用于防止模型过拟合。 12. **迁移学习(Transfer Learning)**:利用在一个任务上训练好的模型来提高另一个相关任务的性能。 深度学习在许多领域都取得了显著的成就,但它也面临着一些挑战,如对大量数据的依赖、模型的解释性差、计算资源消耗大等。研究人员正在不断探索的方法来解决这些问题。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值