课程三(Structuring Machine Learning Projects),第二周(ML strategy(2)) —— 1.Machine learning Flight simulator:Autonomous driving (case study)




【中文翻译】
为了帮助您练习机器学习的策略, 在本周我们将介绍另一个场景, 并询问您将如何行动。我们认为, 这个工作在一个机器学习项目的 "模拟器" 将给一个任务, 告诉你一个机器学习项目像什么!
你受雇于一自动驾驶汽车公司。您负责检测图像中的路标 (停车标志、行人过路标志、建筑前方标志) 和交通信号灯 (红色和绿色灯)。目标是识别这些对象中的哪一个出现在每个图像中。举例来说, 上述图则载有行人过路标志及红色交通灯。
你刚刚开始这个项目。你做的第一件事是什么,假设下面的每个步骤都需要花费相当的时间 (几天)?(B)

B、假
【解释】
1)Softmax 将是一个很好的选择, 如果一个和唯一的可能性 (停止标志, 速度颠簸, 行人过路, 绿灯和红灯) 是存在于每一个图像。
2)多任务学习(Multi-task learning)与softmax回归(Softmax regression)的主要区别在于,softmax将单个标签分配给单个样本,而多任务学习可能将多个标签分给单个样本。
如果输出ŷ是个4维向量:
(1)对于多任务学习,代价函数的计算方式为:
(2) 对于softmax回归,典型的损失计算 一般都会使用:




【解释】
Focus on images that the algorithm got wrong. Also, 500 is enough to give you a good initial sense of the error statistics. There’s probably no need to look at 10,000, which will take a long time.


【解释】
As seen in the lecture on multi-task learning, you can compute the cost such that it is not influenced by the fact that some entries haven’t been labeled.



【解释】
Yes. As seen in lecture, it is important that your dev and test set have the closest possible distribution to “real”-data. It is also important for the training set to contain enough “real”-data to avoid having a data-mismatch problem.






【解释】
The algorithm does better on the distribution of data it trained on. But you don’t know if it’s because it trained on that no distribution or if it really is easier. To get a better sense, measure human-level error separately on both distributions.





【解释】
Yes. You will probably not improve performance by more than 2.2% by solving the raindrops problem. If your dataset was infinitely big, 2.2% would be a perfect estimate of the improvement you can achieve by purchasing a specially designed windshield wiper that removes the raindrops.
是的。通过解决雨滴问题, 您可能不会提高超过2.2% 的性能。如果你的数据集是无限大, 你通过购买一个专门设计的挡风玻璃雨刷, 消除雨滴的方式来获取图片,2.2% 将是一个完美的估计。


【解释】
Yes. If the synthesized images look realistic, then the model will just see them as if you had added useful data to identify road signs and traffic signals in a foggy weather. I will very likely help.
是的。如果合成的图像看起来逼真, 那么模型就会看到它们, 就好像你添加了有用的数据,在大雾天气中来识别路标和交通信号。这很可能会帮忙。
对于 There is little risk of overfitting to the 1,000 pictures of fog so long as you are combing it with a much larger (>>1,000) of clean/non-foggy images.
这句是错误的,注意ittle表示否定含义,整个句子的意识是:只要你用更大的 (>> 1000) 的清楚的/没有雾的图像合成图片, 就不会有过拟合1000张有雾图片的风险,但其实是有风险的。






【解释】
Yes. You have trained your model on a huge dataset, and she has a small dataset. Although your labels are different, the parameters of your model have been trained to recognize many characteristics of road and traffic images which will be useful for her problem. This is a perfect case for transfer learning, she can start with a model with the same architecture as yours, change what is after the last hidden layer and initialize it with your trained parameters.



【解释】
Yes. The problem he is trying to solve is quite different from yours. The different dataset structures make it probably impossible to use transfer learning or multi-task learning.


【解释】
Yes. (A) is an end-to-end approach as it maps directly the input (x) to the output (y).


【解释】
Yes. In many fields, it has been observed that end-to-end learning works better in practice, but requires a large amount of data.
课程三(Structuring Machine Learning Projects),第二周(ML strategy(2)) —— 1.Machine learning Flight simulator:Autonomous driving (case study)的更多相关文章
- 20172313『Java程序设计』课程结对编程练习_四则运算第二周阶段总结
20172313『Java程序设计』课程结对编程练习_四则运算第二周阶段总结 结对伙伴 20172326康皓越 博客地址(http://www.cnblogs.com/326477465-a/p/90 ...
- 20172326『Java程序设计』课程结对编程练习_四则运算第二周阶段总结
20172326『Java程序设计』课程结对编程练习_四则运算第二周阶段总结 小组成员 20172313 余坤澎 20172332 于欣月 20172326 康皓越 小组编程照片 设计思路 通过一个E ...
- 20172311『Java程序设计』课程 结对编程练习_四则运算第二周阶段总结
20172311『Java程序设计』课程 结对编程练习_四则运算第二周阶段总结 结对伙伴 学号 :20172307 姓名 :黄宇瑭 伙伴第一周博客地址 对结对伙伴的评价:黄宇瑭同学的优势在于能够想出一 ...
- 20172321『Java程序设计』课程 结对编程练习_四则运算第二周阶段总结
20172321『Java程序设计』课程 结对编程练习_四则运算第二周阶段总结 结对伙伴 学号 :20172324 姓名 :曾程 伙伴第一周博客地址: 对结对伙伴的评价:一个很优秀的同学,在这次项目中 ...
- 20172325『Java程序设计』课程 结对编程练习_四则运算第二周阶段总结
20172325『Java程序设计』课程 结对编程练习_四则运算第二周阶段总结 结对伙伴 学号:20172306 姓名:刘辰 结对伙伴博客链接 刘辰同学对编程的积极程度很高,并且在编程能力上很不错,有 ...
- 20172332『Java程序设计』课程结对编程练习_四则运算第二周阶段总结
20172313『Java程序设计』课程结对编程练习_四则运算第二周阶段总结 小组成员 20172326康皓越 20172313余坤澎 20172332于欣月 小组编程照片 设计思路 设计一个生成符号 ...
- 课程三(Structuring Machine Learning Projects),第一周(ML strategy(1)) —— 1.Machine learning Flight simulator:Bird recognition in the city of Peacetopia (case study)
[]To help you practice strategies for machine learning, the following exercise will present an in-de ...
- 课程一(Neural Networks and Deep Learning),第二周(Basics of Neural Network programming)—— 4、Logistic Regression with a Neural Network mindset
Logistic Regression with a Neural Network mindset Welcome to the first (required) programming exerci ...
- 20172302『Java程序设计』课程 结对编程练习_四则运算第二周阶段总结
一.结对对象 姓名:周亚杰 学号:20172302 担任角色:驾驶员(周亚杰) 伙伴第二周博客地址 二.本周内容 (一)继续编写上周未完成代码 1.本周继续编写代码,使代码支持分数类计算 2.相关过程 ...
随机推荐
- [leetcode]282. Expression Add Operators 表达式添加运算符
Given a string that contains only digits 0-9 and a target value, return all possibilities to add bin ...
- [leetcode]34.Find First and Last Position of Element in Sorted Array找区间
Given an array of integers nums sorted in ascending order, find the starting and ending position of ...
- node.js中使用zlib模块进行数据压缩和解压
我们可以使用 zlib 模块来对数据进行压缩和解压处理,减小数据体积,加快传输速度. 一.通过创建转换流,对文件进行压缩和解压 const fs = require('fs'); const zlib ...
- playframework 一步一步来 之 日志(一)
日志模块是一个系统中必不可少的一部分,它可以帮助我们写程序的时候查看错误信息,利于调试和维护,在业务面,它也可以记录系统的一些关键性的操作,便于系统信息的监控和追踪. play的日志是基于logbac ...
- E: Sub-process /usr/bin/dpkg returned an error code
E: Sub-process /usr/bin/dpkg returned an error code (1)错误解决 在用apt-get安装软件时出现了类似于install-info: No dir ...
- Python中删除easy_install安装的包
网上查了一大圈,终于在官网上找到了.记一下,备忘...
- [uboot] (第一章)uboot流程——概述
http://blog.csdn.net/ooonebook/article/details/52939100 [uboot] uboot流程系列: [project X] tiny210(s5pv2 ...
- 别人的Linux私房菜(11)认识与学习BASH
Linux下使用BASH Bourne Again Shell 另外一种由用于Unix的伯克利大学的Bill Joy设计的C Shell 系统中合法的shell会写入到/etc/sh ...
- DOM追加笔记
根据 W3C 的 HTML DOM 标准,HTML 文档中的所有内容都是节点: 整个文档是一个文档节点 每个 HTML 元素是元素节点 HTML 元素内的文本是文本节点 每个 HTML 属性是属性节点 ...
- 将JSON数据转换成JAVA的实体类
思路:首先将JSON格式的数据转换成JSONObject,然后将JSONObject转换成Java的实体类(其中类属性包括List等类型) Java实体类: SearchFilter 类 1 publ ...

