Burger - UVa 557 滚动数组 暴力 递推 打表

本文介绍了一个关于概率计算的问题,即在特定条件下计算两个孩子获得相同类型汉堡的概率。通过递推算法和滚动数组技术,解决了这个问题,并给出了AC代码。

Burger 

When Mr. and Mrs. Clinton's twin sons Ben and Bill had their tenth birthday, the party was held at the McDonald's restaurant at South Broadway 202, New York. There were 20 kids at the party, including Ben and Bill. Ronald McDonald had made 10 hamburgers and 10 cheeseburgers and when he served the kids he started with the girl directly sitting left of Bill. Ben was sitting to the right of Bill. Ronald flipped a (fair) coin to decide if the girl should have a hamburger or a cheeseburger, head for hamburger, tail for cheeseburger. He repeated this procedure with all the other 17 kids before serving Ben and Bill last. Though, when coming to Ben he didn't have to flip the coin anymore because there were no cheeseburgers left, only 2 hamburgers.


Ronald McDonald was quite surprised this happened, so he would like to know what the probability is of this kind of events. Calculate the probability that Ben and Bill will get the same type of burger using the procedure described above. Ronald McDonald always grills the same number of hamburgers and cheeseburgers.

Input 

The first line of the input-file contains the number of problems n , followed by n times:


a line with an even number [2,4,6,...,100000], which indicates the number of guests present at the party including Ben and Bill.

Output 

The output consists of n lines with on each line the probability (4 decimals precise) that Ben and Bill get the same type of burger.


Note: a variance of $\pm 0.0001$ is allowed in the output due to rounding differences.

Sample Input 

3
6
10
256

Sample Output 

0.6250
0.7266
0.9500

题意:找到掷n-2次硬币中,同一面至少掷n/2次的概率。

思路:递推,转移方程很好理解,另外用到了滚动数组,这样就可以递推到100000,然后打表解决10000之后的。我打表的时候答案精度是0.00003,所以没用的数可能比较多,正负误差在0.0001即可。

AC代码如下:

#include<cstdio>
#include<cstring>
using namespace std;
double dp[2][100010],ans[100010],p=0,f2[92]={0.9920,0.9921,0.9921,0.9922,0.9923,0.9923,0.9924,0.9924,0.9925,0.9926,0.9926,0.9927,0.9927,0.9928,0.9929,0.9929,0.9930,0.9930,0.9931,0.9932,0.9932,0.9933,0.9933,0.9934,0.9935,0.9935,0.9936,0.9936,0.9937,0.9938,0.9938,0.9939,0.9940,0.9940,0.9941,0.9941,0.9942,0.9943,0.9943,0.9944,0.9944,0.9945,0.9946,0.9946,0.9947,0.9947,0.9948,0.9949,0.9949,0.9950,0.9950,0.9951,0.9952,0.9952,0.9953,0.9953,0.9954,0.9955,0.9955,0.9956,0.9956,0.9957,0.9958,0.9958,0.9959,0.9959,0.9960,0.9961,0.9961,0.9962,0.9962,0.9963,0.9964,0.9964,0.9965,0.9965,0.9966,0.9967,0.9967,0.9968,0.9968,0.9969,0.9970,0.9970,0.9971,0.9971,0.9972,0.9973,0.9973,0.9974,0.9974,0.9975};
int f[92]={10000,10154,10310,10470,10634,10802,10974,11150,11330,11514,11704,11898,12096,12300,12508,12722,12942,13168,13400,13638,13882,14132,14390,14654,14926,15206,15492,15788,16092,16404,16726,17058,17400,17752,18114,18488,18872,19268,19678,20100,20536,20986,21452,21934,22432,22946,23478,24030,24602,25194,25808,26444,27104,27788,28498,29236,30004,30802,31632,32496,33396,34334,35312,36332,37398,38510,39674,40890,42164,43498,44896,46362,47902,49520,51220,53010,54896,56884,58982,61198,63540,66020,68648,71436,74398,77548,80902,84478,88296,92380,96754,100000};

int main()
{ int t,n,i,j,k,a,b,num=0;
  dp[1][0]=0.25;
  dp[1][1]=0.5;
  dp[1][2]=0.25;
  ans[2]=0;
  ans[4]=0.25;
  for(i=4;i<=10000;)
  { if((i/2)%2)
    { a=1;b=0;}
    else
    { a=0;b=1;}
    dp[a][0]=dp[b][0]*0.25;
    dp[a][1]=dp[b][0]*0.5+dp[b][1]*0.25;
    for(j=2;j<=i;j++)
     dp[a][j]=dp[b][j-1]*0.5+(dp[b][j-2]+dp[b][j])*0.25;
    i+=2;
    for(j=i/2;j<=i;j++)
     ans[i]+=dp[a][j];
  }
  scanf("%d",&t);
  while(t--)
  { scanf("%d",&n);
    if(n<=10000)
    printf("%.4f\n",ans[n]*2);
    else
    { for(i=0;i<92;i++)
       if(n<=f[i])
       { printf("%.4f\n",f2[i]);
         break;
       }
    }
  }
}




内容概要:本文介绍了基于贝叶斯优化的CNN-LSTM混合神经网络在时间序列预测中的应用,并提供了完整的Matlab代码实现。该模型结合了卷积神经网络(CNN)在特征提取方面的优势与长短期记忆网络(LSTM)在处理时序依赖问题上的强大能力,形成一种高效的混合预测架构。通过贝叶斯优化算法自动调参,提升了模型的预测精度与泛化能力,适用于风电、光伏、负荷、交通流等多种复杂非线性系统的预测任务。文中还展示了模型训练流程、参数优化机制及实际预测效果分析,突出其在科研与工程应用中的实用性。; 适合人群:具备一定机器学习基基于贝叶斯优化CNN-LSTM混合神经网络预测(Matlab代码实现)础和Matlab编程经验的高校研究生、科研人员及从事预测建模的工程技术人员,尤其适合关注深度学习与智能优化算法结合应用的研究者。; 使用场景及目标:①解决各类时间序列预测问题,如能源出力预测、电力负荷预测、环境数据预测等;②学习如何将CNN-LSTM模型与贝叶斯优化相结合,提升模型性能;③掌握Matlab环境下深度学习模型搭建与超参数自动优化的技术路线。; 阅读建议:建议读者结合提供的Matlab代码进行实践操作,重点关注贝叶斯优化模块与混合神经网络结构的设计逻辑,通过调整数据集和参数加深对模型工作机制的理解,同时可将其框架迁移至其他预测场景中验证效果。
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