CS 5436 / INFO 5303

「文智助手」全场景办公AI助理AI+OPC全流程实践方案(大赛作品) 阅读详情

CS 5436 / INFO 5303 Fall 2024

Homework 1

Due: September 24, 11:59p ET

This is an INDIVIDUAL assignment.

You may discuss, but each student must submit their own work.

White space after each problem indicates the approximate length of the expected answer.

REQUIRED (40 points)

Problem 1 (6 points)

The same-origin policy in Web browsers is supposed to prevent information flows between Web content from different origins.  

(a) Why and how does it allow third-party trackers to learn about first-party URLs visited by the browser?

(b) How can the same-origin policy be strengthened to prevent tracking via third-party cookies?  Mention any concrete examples that we discussed in class.  (Note: Google’s Privacy Sandbox does not involve a redesign of the same-origin policy).

(c) What forms of tracking would still work even if the browser completely blocks all information flows between Web origins?

Problem 2 (2 points)

In some cases, sharing personally identifiable information in a raw form. can be illegal, but sharing hashed versions is allowed. Discuss the differences between raw and hashed identifiers, including the related privacy implications.

Problem 3 (6 points)

In this problem, you will investigate and evaluate several privacy-enhancing technologies.

(a) Describe what Chrome’s Incognito browsing mode provides. What types of tracking techniques are blocked in this mode?  What types of tracking still work?

(b) Cover Your Tracks is a service that allows you to measure the uniqueness of browser fingerprints. What parts of your browser fingerprint provide the most identifying information (please describe your configuration: OS, browser make and model, etc.)?  What can you do to reduce the identifiability of your fingerprint?

(c) Recent research has shown that browser fingerprints change frequently from normal use. Suppose that, on average, the fingerprint of your browser changes every 5 days. To what extent do these changes protect you from tracking, and to what extent do they not?

Problem 4 (2 points)

Compare and contrast Do Not Track and Global Privacy Control.  What are the differences in enforcement?  Do you think GPC will be more successful than DNT and, if so, why?

Problem 5 (2 points)

Compare and contrast Mozilla’s Total Cookie Protection and Google’s Chrome privacy sandbox. Which provides stronger privacy protection?  Why?

Problem 6 (6 points)

(a) How do the main tracking mechanisms differ between Web and mobile?

(b) Why is it easier to identify mobile users?  

(c) What are other key differences between Web and mobile tracking?  

Problem 7 (5 points)

We discussed two categories of privacy definitions: control and secrecy.

(a) Cite two privacy theorists in each of these categories.

(b) Select one from each. Summarize their definitions and explain why they belong in respective categories

(c) How would each of the two theories explain why certain online behavioral tracking violate privacy.  

Problem 8 (3 points)

(a) According to contextual integrity (CI), a right to privacy is a right to _____________?

(b) In terms of CI, defining privacy as a right to control information about oneself is right/wrong?  Explain.

Problem 9 (6 points)

Contextual informational (privacy) norms/rules have five parameters.

(a) What does this mean?

(b) List the parameters, briefly describe each, and illustrate with an example.

(c) What does “acting in capacity” mean?

(d) From your home background (culture) state a well formed contextual informational rule/norm that you think might surprise some of your classmates. NOTE: These need not be legal rules.

(e - 2pt bonus)  In 2009, Amazon deleted copies of George Orwell’s 1984 from customers’ Kindle devices. (Yes, they refunded the 99 cents that customers had paid.) There are many perspectives one could take on this action, but how might you use CI as a lens through which to see the privacy aspect?

Problem 10 (2 points)

In analyzing whether a system or a practice violates privacy, contextual integrity (CI) draws attention to factors that may be overlooked by other definitions. Recall the weather, family-tracking, and cheap-gas mobile apps that share users’ driving information with data brokers, who then derive an individualized driving score for each user and sell these scores to auto insurance companies.  

Use the CI framework to reveal all features of this case that are relevant to the question of whether these apps violate users’ privacy.   (Note: Your judgment on whether they do or do not is a separate question which we will ask in a later homework.)

PICK YOUR OWN – should add up to 20 points

For each problem you pick, do the entire problem (i.e., you cannot choose-and-mix subproblems)

Problem CS1 (5 points)

How do ad-blocking tools like AdBlock Plus, Ghostery, and Disconnect mitigate propagation of personal information to advertisers via RTB?

Problem CS2 (15 points)

OpenWPM is an open-source Web privacy measurement framework.  Use it to analyze 100 popular websites from diverse categories (explain your algorithm for selecting these sites).  

(a - 3 points) What are the most common trackers on your chosen sites?  

(b - 6 points) How many of the trackers include persistent identifiers?  What do you consider a persistent identifier? Identify fingerprinting code (you can use this link for the list of probable fingerprinting commands – try recursive_dump_page_source method).

(c - 6 points) Count the flows of information across different contexts.  To do this, assign each site to a particular social context (eg, news, entertainment, etc.) – it’s okay to use an automated assignment algorithm.  There exists an information flow between context A and context B if the same tracker appears in a page that belongs to context A and a page that belongs to context B.

You can upload a documented Jupyter notebook or report in PDF.

Useful links and helpful hints:

● OpenWPM tutorial

● OpenWPM configuration settings available

● You can create your own function in custom_command.py (there’s already an example function there)

○ Might be useful: Selenium with Python for programmatically accessing webpage elements (this is what the LinkCountingCommand() there does)

● openwpm/command_sequence.py might also have some useful stuff

○ command_sequence.recursive_dump_page_source() will dump the page source to the sources directory

● crawl-data.sqlite contains some information retrieved from the web crawl

○ This link explains some of the tables it stores

Problem INFO1 (7 points)

Briefly summarize the three connections between privacy and autonomy discussed in PIC. Make sure you provide a definition of autonomy that serves as the basis for your answer

Problem INFO2 (2 points)

“Privacy is a right, not a preference!” is a slogan popularized by the Electronic Privacy Information Center. What’s the big deal?

Problem INFO3 (1 point)

“All the parameters matter!”  Explain.

Problem INFO4 (3 points)

(a) What is meant by “privacy harms”?

(b) List and briefly describe two, making sure you explain why they are privacy harms.

Problem INFO5 (1 point)

Tom Gerety was concerned that privacy “expands like a gas to fill up available space.”

What was his (and others’) approach to addressing this concern?

Problem INFO6 (6 points)

“If you’ve done nothing wrong (got nothing to hide,) you don’t need privacy!”

(a) Expand on this claim, making the most plausible case in its favor.  Use examples.

(c) How might you draw on the definition of Contextual Integrity to push back? Feel free to make up your own examples or draw any from lectures that are useful for your answer.

船载 OpenHarmony AIS 航海工作台:告警与防碰撞预警系统 船载OpenHarmony AIS航海工作台的告警与防碰撞系统,集成十余类告警(超速、偏航、会遇、SOS等),通过网格空间索引实现高效邻近判定,保障实时性。告警管理器统一调度规则、去重、通知与日志,支持悬浮球、语音、跑马灯、弹窗多渠道触达,确保航行安全。SOS紧急通道独立运行,避免干扰。系统设计强调性能优化与多场景适配,为航海安全提供核心支撑。 阅读详情

相关推荐

AI 应用开发:四维架构落地实战——以 HR Agent 为例拆解感知·推理·记忆·执行【意图路由、规划编排、长期记忆、工具执行、面试题】

本文基于四维架构(感知·推理·记忆·执行),以HR Agent为例,系统拆解大模型应用从“聊天”到“生产可用”的落地路径。通过Spring AI实现意图路由、规划编排、RAG检索与工具调用,结合真实踩坑复盘,揭示架构化设计对安全、成本与可维护性的关键价值,为AI Agent工程化提供可复用的实战框架。

王大师企业官方博客 1323

CS 5436 / INFO 5303 Fall 2024 Homework 1Python

yourWhy?eachprivacy。

wx___codinghelp的博客 1080

cs5463中文资料

概述 CS5463 是一个包含两个ΔΣ模-数转换器(ADC)、功率计算功能、电能到频率转换器和一个串行接口的 完整的功率测量芯片。它可以精确测量瞬时电压,电流和计算IRMS、VRMS、瞬时功率、有功功率、无功功率,用 于研制开发单相、2线或3线电表。 CS5463可以使用低成本的分流器或互感器测量电流,使用分压电阻或电压互感器测量电压。 CS5463具有与微控制器通讯的双向串口,可编程的电能-脉冲输出功能。CS5463还具有方便的片上系统校准 功能。具有温度传感器,电压下降检测,相位补偿功能。

AI 问数真的能替代 BI 报表吗?

AI问数赋能自然语言查询,提升临时分析效率,但无法替代传统报表。报表适用于稳定输出与持续监控,AI则擅长探索性提问。企业用好AI问数,关键在于夯实数据模型、统一指标口径、健全权限体系与结果校验机制。合理组合报表、仪表板与AI,才能实现高效、准确的业务洞察。

葡萄城技术团队博客 339

从零开始用TS开发babylonjs - 本机环境搭建

babylonjs,初始化本地安装,web3d引擎

han3527638的专栏 798

Android - 蓝牙开发

文章目录科普SIG类型制式编程Classic BluetoothBluetooth Low Energy蓝牙的UUID是什么?有什么用? 科普 https://zh.wikipedia.org/wiki/%E8%97%8D%E7%89%99 蓝牙(英语:Bluetooth),一种无线通讯技术标准,用来让固定与移动设备,在短距离间交换资料,以形成个人局域网(PAN)。其使用短波特高频(UHF)无线电波,经由 2.4 至 2.485 GHz 的 ISM(工业、科学、医疗)频段来进行通信。1994 年由电信商

qq_43323826的博客 1万+

ChatGPT、Perplexity、Gemini 与 AI Overviews 引用对比:谁的引用更可信

AI 引擎的引用机制差异很大:ChatGPT 以行内引用加 Sources 列表呈现来源,Perplexity 几乎逐句可溯源,Gemini 在回答中内嵌相关来源链接,AI Overviews 则把多源聚合答案与来源列表一起放在搜索结果顶部。本文从引用透明度、来源偏好、可干预性 3 个维度对比这 4 个引擎,并给出 GEO 的 3 点启示,帮企业在多引擎环境下提升被引用的概率。

淡然一笑 的专栏 276

实例讲解Simulink应用层开发CAN报文解包及CAN信号设置方法

本文通过讲解MCU CAN解包模块的模块设置、输入输出信号的设置及仿真测试验证,说明了CAN解包模块的设置方法、CAN解包模块输入输出信号的设置方法以及CAN解包模块的仿真测试及修正方法。在模型搭建过程中,建议养成一种搭建完成仿真测试验证的习惯,在模块阶段发现错误,解决错误,避免在后期模块集成之后报出大量错误,增加解决错误的工作量和难度。希望通过本文的介绍能给相关技术人员带来一些参考和帮助。

qq_43011586的博客 5684

【从零理解 Spring AI · 12】MCP 的两种角色:做客户端还是做服务端

上一篇讲了 MCP 的概念:标准协议、三大角色、M+N 替代 M×N。

nice_lcj520的博客 615

JSP期末考试复习习题及答案

以下习题是对于平时课堂小测的整理,如有错误敬请指正! 欢迎一起交流~ 预祝各位小伙伴们考试顺利! Test1 一. 单选题 – 1.(单选题)当用户请求jsp页面时,JSP引擎就会执行该页面的字节码文件响应客户的请求,执行字节码文件的结果是() A. 发送一个JSP源文件到客户端 B. 发送一个Java文件到客户端 C. 发送一个HTML页面到客户端 D. 什么都不做 – 2.(单选题)下列动态网页和静态网页的根本区别描述错误的是( D) A. 静态网页服务器端返回的HTML文件是事先存储好的 B. 动.

技术小白 2万+

从零到一:手把手教你用SAM-BA工具完成嵌入式系统烧录

本文详细介绍了如何使用SAM-BA工具完成嵌入式系统烧录的全过程,包括环境搭建、硬件连接、Bootloader烧录以及完整系统镜像的烧录步骤。针对常见问题提供了实用的排查指南,并分享了批量烧录和自定义脚本的高级技巧,帮助开发者高效完成嵌入式系统开发。

weixin_42523792的博客 382

抽象主义人工智能和生命体AI的内在联系

彭军辉提出“抽象主义人工智能”作为第四大AI范式,批判符号主义、连接主义、行为主义剥离生命本体的误区,确立“智能是现象,生命是本体”的核心公理。主张原生智能必须依附具备存续诉求的生命本体,提出“生命体AI理论”,构建七层演化阶梯与自闭环工程框架,强调智能不可跳过底层存续机制直接模拟语言。该体系为生成式AI提供科学边界,推动从“仿真表象”向“原生智能”的范式跃迁,开辟硅基人工生命新赛道。

抽象主义人工智能和生命体AI 207

DBeaver 实战:手动配置 PostgreSQL JDBC 驱动的高级技巧

本文详细介绍了在DBeaver中手动配置PostgreSQL JDBC驱动的高级技巧。针对企业内网、多版本数据库兼容等场景,从获取官方驱动文件、创建自定义驱动配置,到使用Maven坐标管理、连接参数调优及解决常见连接问题,提供了完整的实战指南,帮助开发者实现对数据库连接环境的精准掌控。

apple5的专栏 837

OpenAI Control Room 架构拆解:给企业 AI Agent 加一套可审计的急停系统

OpenAI 公开了用于追踪和停止研究模型工作负载的 Control Room。本文拆解 job_id、推理前鉴权、35 秒撤销窗口与 Fail Closed 设计,并给出企业 AI Agent 急停控制面的数据模型、FastAPI 参考实现、缓存撤销策略和故障演练清单。

jerry764的博客 343

IGBT基础知识及应用电路仿真(Multisim)

本文系统介绍了IGBT(绝缘栅双极晶体管)及其应用电路。首先详细解析了IGBT的基本结构、工作原理和特性参数,重点阐述了其五层半导体结构、电导调制效应和电压驱动特性。随后通过Multisim仿真展示了IGBT在四种典型功率变换电路中的应用:半桥驱动电路实现隔离式互补驱动;全桥逆变电路完成DC-AC转换;三相桥式逆变电路生成三相交流电;直流斩波电路(BUCK/BOOST/BUCK-BOOST)实现电压升降变换。文章通过理论分析与仿真波形相结合,全面呈现了IGBT在中高压大功率场景下的优异性能和灵活应用。

XU157303764的博客 1221

【语音分析】语音倒谱分析(变声器 男变女 女变男 童声)【含GUI Matlab源码 4247期】

语音倒谱分析(变声器 男变女 女变男 童声) 完整的代码,方可运行;可提供运行操作视频!适合小白!

订阅付费专栏Matlab(奶茶价版),可赠送奶茶价版付费专栏指定代码1份; 828

OpenClaw 深度调查:把 AI 助手做成你自己的控制平面

对 openclaw/openclaw 的完整技术尽调:Gateway 架构、Agent 运行时、工具与沙箱、30 个消息通道、181 个插件、安全信任边界、工程代价与采用建议。17 章 / 10 张图,全部数字标注口径,数据截至 2026-10-31。

厚积薄发 349

ET2012新版免狗免安装绿色软件

ET2012新版免狗免安祖昂绿色软件。,新界面,新视觉,

AI赋能智慧港口建设|造形家快速生成港口全要素三维场景

随着智慧港口、港口数字孪生项目的持续推进,港口三维场景的需求日益旺盛。传统港口三维建模,往往需要航拍倾斜摄影采集数据,再人工搭建吊机、岸桥、船舶、集装箱等大量工业设施模型,项目周期长、人力成本高,修改迭代繁琐,制约港口可视化项目落地效率。国产AI三维建模工具,依托卫星影像 AI 识别能力,无需航拍作业,一键快速搭建港口完整三维场景,为港口数字孪生、方案汇报、投标演示提供高效的场景解决方案。

m0_69539167的博客 86

visio安装包

Microsoft Office Visio 是一款专业的制图软件。利用经过更新的形状、协作工具和数据链接图表,创建专业级图表以简化复杂的信息。

jre9下载 jdk9

(Java SE Runtime Environment 9.0) 9.0u175 官方最新版 【x86|x64】

从设备增点到治理闭环:老旧小区物业数智化落地实践

系统连接合同审批、银行流水、收费记录与项目台账,执行到期提醒、收入核验、异常标色和月度对账,让每一笔资金来源明确、去向可查。针对设施老旧、诉求响应慢、公共收益不透明等问题,津南以6类小区为试点场景,统一建设物业数字化底座,关联楼栋、单元、房屋、业主、空间、设施设备、工单、公共收益等核心数据。系统采用统一编码管理,将摄像头、路灯、井盖、道路、绿化等设施纳入资产台账,解决纸质记录更新滞后、故障责任难追溯的问题。基于历史工单建立分类模型和知识库,辅助识别照明、给排水、电梯、环境秩序、秩序维护等问题;

wuyeshuzhihua的博客 161

深度学习工具包 DeepLearningToolbox

深度学习工具包Deprecation notice.-----This toolbox is outdated and no longer maintained.There are much better tools available for deep learning than this toolbox, e.g. [Theano](http://deeplearning.net/software/theano/), [torch](http://torch.ch/) or [tensorflow](http://www.tensorflow.org/)I would suggest you use one of the tools mentioned above rather than use this toolbox.Best, Rasmus.DeepLearnToolbox================A Matlab toolbox for Deep Learning.Deep Learning is a new subfield of machine learning that focuses on learning deep hierarchical models of data.It is inspired by the human brain's apparent deep (layered, hierarchical) architecture.A good overview of the theory of Deep Learning theory is[Learning Deep Architectures for AI](http://www.iro.umontreal.ca/~bengioy/papers/ftml_book.pdf)For a more informal introduction, see the following videos by Geoffrey Hinton and Andrew Ng.* [The Next Generation of Neural Networks](http://www.youtube.com/watch?v=AyzOUbkUf3M) (Hinton, 2007)* [Recent Developments in Deep Learning](http://www.youtube.com/watch?v=VdIURAu1-aU) (Hinton, 2010)* [Unsupervised Feature Learning and Deep Learning](http://www.youtube.com/watch?v=ZmNOAtZIgIk) (Ng, 2011)If you use this toolbox in your research please cite [Prediction as a candidate for learning deep hierarchical models of data](http://www2.imm.dtu.dk/pubdb/views/publication_details.php?id=6284)```@MASTERSTHESIS\{IMM2012-06284, author = "R. B. Palm", title = "Prediction as a candidate for learning deep hierarchical models of data", year = "2012",}```Contact: rasmusbergpalm at gmail dot comDirectories included in the toolbox-----------------------------------`NN/` - A library for Feedforward Backpropagation Neural Networks`CNN/` - A library for Convolutional Neural Networks`DBN/` - A library for Deep Belief Networks`SAE/` - A library for Stacked Auto-Encoders`CAE/` - A library for Convolutional Auto-Encoders`util/` - Utility functions used by the libraries`data/` - Data used by the examples`tests/` - unit tests to verify toolbox is workingFor references on each library check REFS.mdSetup-----1. Download.2. addpath(genpath('DeepLearnToolbox'));Example: Deep Belief Network---------------------```matlabfunction test_example_DBNload mnist_uint8;train_x = double(train_x) / 255;test_x = double(test_x) / 255;train_y = double(train_y);test_y = double(test_y);%% ex1 train a 100 hidden unit RBM and visualize its weightsrand('state',0)dbn.sizes = [100];opts.numepochs = 1;opts.batchsize = 100;opts.momentum = 0;opts.alpha = 1;dbn = dbnsetup(dbn, train_x, opts);dbn = dbntrain(dbn, train_x, opts);figure; visualize(dbn.rbm{1}.W'); % Visualize the RBM weights%% ex2 train a 100-100 hidden unit DBN and use its weights to initialize a NNrand('state',0)%train dbndbn.sizes = [100 100];opts.numepochs = 1;opts.batchsize = 100;opts.momentum = 0;opts.alpha = 1;dbn = dbnsetup(dbn, train_x, opts);dbn = dbntrain(dbn, train_x, opts);%unfold dbn to nnnn = dbnunfoldtonn(dbn, 10);nn.activation_function = 'sigm';%train nnopts.numepochs = 1;opts.batchsize = 100;nn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.10, 'Too big error');```Example: Stacked Auto-Encoders---------------------```matlabfunction test_example_SAEload mnist_uint8;train_x = double(train_x)/255;test_x = double(test_x)/255;train_y = double(train_y);test_y = double(test_y);%% ex1 train a 100 hidden unit SDAE and use it to initialize a FFNN% Setup and train a stacked denoising autoencoder (SDAE)rand('state',0)sae = saesetup([784 100]);sae.ae{1}.activation_function = 'sigm';sae.ae{1}.learningRate = 1;sae.ae{1}.inputZeroMaskedFraction = 0.5;opts.numepochs = 1;opts.batchsize = 100;sae = saetrain(sae, train_x, opts);visualize(sae.ae{1}.W{1}(:,2:end)')% Use the SDAE to initialize a FFNNnn = nnsetup([784 100 10]);nn.activation_function = 'sigm';nn.learningRate = 1;nn.W{1} = sae.ae{1}.W{1};% Train the FFNNopts.numepochs = 1;opts.batchsize = 100;nn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.16, 'Too big error');```Example: Convolutional Neural Nets---------------------```matlabfunction test_example_CNNload mnist_uint8;train_x = double(reshape(train_x',28,28,60000))/255;test_x = double(reshape(test_x',28,28,10000))/255;train_y = double(train_y');test_y = double(test_y');%% ex1 Train a 6c-2s-12c-2s Convolutional neural network %will run 1 epoch in about 200 second and get around 11% error. %With 100 epochs you'll get around 1.2% errorrand('state',0)cnn.layers = { struct('type', 'i') %input layer struct('type', 'c', 'outputmaps', 6, 'kernelsize', 5) %convolution layer struct('type', 's', 'scale', 2) %sub sampling layer struct('type', 'c', 'outputmaps', 12, 'kernelsize', 5) %convolution layer struct('type', 's', 'scale', 2) %subsampling layer};cnn = cnnsetup(cnn, train_x, train_y);opts.alpha = 1;opts.batchsize = 50;opts.numepochs = 1;cnn = cnntrain(cnn, train_x, train_y, opts);[er, bad] = cnntest(cnn, test_x, test_y);%plot mean squared errorfigure; plot(cnn.rL);assert(er<0.12, 'Too big error');```Example: Neural Networks---------------------```matlabfunction test_example_NNload mnist_uint8;train_x = double(train_x) / 255;test_x = double(test_x) / 255;train_y = double(train_y);test_y = double(test_y);% normalize[train_x, mu, sigma] = zscore(train_x);test_x = normalize(test_x, mu, sigma);%% ex1 vanilla neural netrand('state',0)nn = nnsetup([784 100 10]);opts.numepochs = 1; % Number of full sweeps through dataopts.batchsize = 100; % Take a mean gradient step over this many samples[nn, L] = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.08, 'Too big error');%% ex2 neural net with L2 weight decayrand('state',0)nn = nnsetup([784 100 10]);nn.weightPenaltyL2 = 1e-4; % L2 weight decayopts.numepochs = 1; % Number of full sweeps through dataopts.batchsize = 100; % Take a mean gradient step over this many samplesnn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');%% ex3 neural net with dropoutrand('state',0)nn = nnsetup([784 100 10]);nn.dropoutFraction = 0.5; % Dropout fraction opts.numepochs = 1; % Number of full sweeps through dataopts.batchsize = 100; % Take a mean gradient step over this many samplesnn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');%% ex4 neural net with sigmoid activation functionrand('state',0)nn = nnsetup([784 100 10]);nn.activation_function = 'sigm'; % Sigmoid activation functionnn.learningRate = 1; % Sigm require a lower learning rateopts.numepochs = 1; % Number of full sweeps through dataopts.batchsize = 100; % Take a mean gradient step over this many samplesnn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');%% ex5 plotting functionalityrand('state',0)nn = nnsetup([784 20 10]);opts.numepochs = 5; % Number of full sweeps through datann.output = 'softmax'; % use softmax outputopts.batchsize = 1000; % Take a mean gradient step over this many samplesopts.plot = 1; % enable plottingnn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');%% ex6 neural net with sigmoid activation and plotting of validation and training error% split training data into training and validation datavx = train_x(1:10000,:);tx = train_x(10001:end,:);vy = train_y(1:10000,:);ty = train_y(10001:end,:);rand('state',0)nn = nnsetup([784 20 10]); nn.output = 'softmax'; % use softmax outputopts.numepochs = 5; % Number of full sweeps through dataopts.batchsize = 1000; % Take a mean gradient step over this many samplesopts.plot = 1; % enable plottingnn = nntrain(nn, tx, ty, opts, vx, vy); % nntrain takes validation set as last two arguments (optionally)[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');```[![Bitdeli Badge](https://d2weczhvl823v0.cloudfront.net/rasmusbergpalm/deeplearntoolbox/trend.png)](https://bitdeli.com/free "Bitdeli Badge")

企业微信iPad协议详解:接口接入、实时消息同步与语音转码实践

深入解析企业微信iPad协议的技术架构与开发思路,涵盖API接口接入、长连接通信、实时消息同步、Silk语音转码、异常重连及高并发优化,帮助开发者了解企业微信自动化系统的设计与实现。

欢迎来到祖安! 513

Product Hunt 热榜 2026-11-13:Plurai Simulation、YouArt Desktop App、Omnigraph

2026-11-13 Product Hunt 新产品精选:Plurai Simulation、YouArt Desktop App、Omnigraph。汇总 15 款产品的中文介绍、功能关键词和产品链接。

LaughingZhu的博客 617

【AI Agent面试系列】AI Agent可观测、评测、安全相关的知识点

之前的文章介绍了和,本文主要介绍AI Agent的测评、可观测等知识。

406

Docling 原理解读:是什么、为什么、怎么样

从论文视角拆解 Docling 的整体架构、DoclingDocument 数据模型、PDF 预处理、版面分析、OCR、TableFormer 双解码器、阅读顺序与装配流程,并解释版面框与 OCR 框的区别。

PLANTTHESON的博客 401

WorkBuddy 实战:一句话需求 → 七屏可交互原型,户外徒步 App 的 AI 设计全流程

WorkBuddy 实战:一句话需求 → 七屏可交互原型,户外徒步 App 的 AI 设计全流程,跟着博主一起来还原整个过程。

夜郎King的博客 5164

从认识人工智能到理解机器学习:我的 AI 入门第一课

正式开始学习人工智能之前,我对 AI 的认识其实比较模糊:知道 ChatGPT、知道大模型,也知道现在很多岗位都在谈 AI,但如果有人问我“人工智能到底是什么”“机器是怎么学会识别一只猫的”,我未必能从头到尾讲清楚。上午的课程更偏向行业全景:人工智能的发展、研究方向、NLP、岗位分工、学习路线和 AI Coding 工具。下午的课程更偏向技术基础:机器学习的原理、有监督学习、深度学习、模型训练和评估方法。

chh563的博客 568

企业宣传片发前如何批量排查并遮挡视频里的敏感信息

做企业视频的人都会碰到一个具体问题:实拍画面里混进了不该出现的信息。办公室背景里有贴着客户信息的白板,采访对象的授权书边缘露出一串身份证号码,街采素材里路人面孔和车牌号直接入镜。人眼在时间线上扫视,会漏掉快速闪过的帧、画面边缘的小字、以及暗光区域里的证件号码。用抹影AI完成车牌、人脸和证件区域批量检测后,又手动补充了两个时间段的屏幕遮罩,最终导出时画面已无合规隐患。重点关注快速运动、遮挡、低光照条件下的目标,这些是漏检的高发区域。导入视频后选择要检测的目标类型(人脸、车牌、证件号、二维码等)。

Mysssone的博客 181
上一篇: COSC2391 Further Programming
下一篇: EMCH792 – Optimal State Estimation
vx___codinghelp
博客等级 码龄2年 778粉丝 101原创
评论
成就一亿技术人!
拼手气红包6.0元
还能输入1000个字符
 
 条评论被折叠 查看
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付:10.00元
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
实付元
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