计算机在线考试摘要,基于WEB的网络在线考试系统-毕业论文中文摘要题目(可编辑).doc...

基于WEB的网络在线考试系统-毕业论文中文摘要题目(可编辑)

基于WEB的网络在线考试系统-毕业论文中文摘要题目

Foreign language Summary of Graduation thesis

东北电力学院 2005 届本科毕业设计(论文) 毕业论文外文摘要

Title: Based on the online examination system of cross-domain of WEB

Summary:

This article has explained a system structure based on examadministrative system of WEB and designs’s and realizes’s process.

Three layers of system structure that the system adopts:

JSP-JavaBean-Sql Server 2000 develops, and with namelying bow is

realized the show logic by JSP's page, and the vocational work logic

complicated inside the package realization system of JavaBean is used

in the centre, and Sql Server's 2000 data bases realizations data

logics are used to the back endThe System puts forward the complete settlement scheme of online

examination system of cross-domain communication, and realized the

classified special principle of title: the fraction etc are had

something made to order by proposition person oneself degree of

difficulty examination questions type examination title. The system

can accomplish having something made to order of examination paper,

the voluntarily formation of examination paper, and can look at in

advance to the examination paper, and revise. The examination paper

issues when accomplishing by the control center. Supervises center

control at candidate for an entrance examination's examination course

by the examination, also can be to candidate for an entrance

examination the examination state is prohibited or is permitted

controls, the examination end is given a mark, and the system can be

according to the answer circumstances of candidate for an entrance

examination to a fraction. The system adopts B / S's model design,

and can add up the functions such as analysis etc to the frequency

that the examination questions were drawn. The system developed

successfully under the terrace of 2000 servers Windows Keywords: Question Management; Exam On-line ; JSP ; JavaBean; Sql

Server 2000 1东北电力学院 2005 届本科毕业设计(论文)

第一章 绪论

1.1 课题背景与意义

1.1.1 课题开发背

非常抱歉,我之前提供的代码存在错误。在 PyTorch 中,并没有直接提供离散余弦变换(DCT)的函数。对于 DCT 的实现,你可以使用 `torch.rfft` 函数结合 DCT 系数矩阵来进行计算。 下面是一个修正后的示例代码: ```python import torch import torch.nn as nn import torch.optim as optim # 定义离散余弦变换(DCT)系数矩阵 dct_matrix = torch.zeros(256, 256) for i in range(256): for j in range(256): dct_matrix[i, j] = torch.cos((2 * i + 1) * j * 3.14159 / (2 * 256)) # 定义 OMP 算法 def omp(A, y, k): m, n = A.shape x = torch.zeros(n, 1) residual = y.clone() support = [] for _ in range(k): projections = torch.abs(A.t().matmul(residual)) index = torch.argmax(projections) support.append(index) AtA_inv = torch.linalg.inv(A[:, support].t().matmul(A[:, support])) x_new = AtA_inv.matmul(A[:, support].t()).matmul(y) residual = y - A[:, support].matmul(x_new) x[support] = x_new return x # 加载原始图像 image = torch.randn(256, 256) # 压缩感知成像 measurement_matrix = torch.fft.fft(torch.eye(256), dim=0).real compressed = measurement_matrix.matmul(image.flatten().unsqueeze(1)) # 使用 OMP 进行重构 reconstructed = omp(dct_matrix, compressed, k=100) # 计算重构误差 mse = nn.MSELoss() reconstruction_error = mse(image, reconstructed.reshape(image.shape)) print("重构误差:", reconstruction_error.item()) ``` 在这个示例中,我们手动定义了 DCT 系数矩阵 `dct_matrix`,然后使用 `torch.fft.fft` 函数计算测量矩阵,并进行实部提取。接下来的步骤与之前的示例相同。 请注意,这只是一个示例,用于演示如何使用自定义的 DCT 系数矩阵进行压缩感知成像。在实际应用中,你可能需要根据具体的需求进行调整和优化。
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