Lecture1-3Applications of ML

最新推荐文章于 2023-12-15 15:33:54 发布
最新推荐文章于 2023-12-15 15:33:54 发布
阅读量248 收藏
点赞数
分类专栏： ML(NTU)
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/zypandora/article/details/48381779
版权

Applications of daily needs

Food

  • data: twitter data(words + location)
  • skill: tell food poisoning likeliness of restaurant properly

Clothing

  • data: sales figures + client surveys
  • skill: give good fashion recommendations to clients

Housing

  • data: characteristics of building and their energy load
  • skill: predict energy load of other buildings

Transportation

  • data: some traffic sign images and meanings
  • skill: recognize traffic signs accurately

detailed applications

Education

  • data: students’ records on quizzes on a math tutoring system
  • skill: predict wether a student can give a correct answer to another quiz
A possible ML solution

answer correctly ≈ [recent strength of student > <script type="math/tex" id="MathJax-Element-16">></script> difficulty of question]

  • give ML 9 million records from 3000 students
  • ML determines strength and difficulty automatically

Entertainment: Recommender System

  • data: how many users have rated some movies
  • skill: how a user would rate an unrated movie
A possible ML solution

这里写图片描述

  • pattern: rating <- viewer/movie factors
  • learning:
    knowing rate -> learned factors -> unknown rating prediction
zypandora
关注
专栏目录
机器学习基石笔记 Lecture 3 - Types of Learning
Note of Transposition
12-19 1176
Lecture 3 - Types of LearningLearning with Different Output Space
Machine Learning Lecture Notes
weixin_43200831的博客
02-02 1422
Machine Learning Note1[Financial Mathematics]
Stanford ML - Lecture 4 - Neural Networks: Representation
D. Qiu的专栏
03-13 858
1. Non-linear hypotheses why introduce non-linear hypotheses? high dimensional datanon-linear hypotheses 2. Neurons and the brain Neural networks origins: algorithms that try to m
Course3-Week2-推荐系统
最新发布
虎慕的博客
12-15 1902
推荐机制、协同过滤算法、基于内容过滤、PCA主成分分析
机器学习基石 Lecture1: The Learning Problem
无所知的博客
11-06 437
机器学习基石 lecture1: The Learning ProblemCourse IntroductionCourse DesignWhat is Machine Learning定义Key Essence of Machine LearningApplications of Machine LearningComponents of LearningMachine Learning and...
斯坦福课程Knowledge Graphs-What is a Knowledge Graph?
anqi3776的博客
09-02 1533
斯坦福课程Knowledge Graphs-What is a Knowledge Graph?1. Introduction2. Knowledge Graph Definition3. Recent Applications of Knowledge Graphs3.1 Knowledge Graphs for organizing Knowledge over the Internet3.2 Knowledge Graphs for Data Integration in Enterprises4.
WEB OF SCIENCE 中检索式的编写
热门推荐
weixin_33895516的博客
05-08 2万+
2019独角兽企业重金招聘Python工程师标准>>> ...
个人网页、博客、课程--不断更新
qq_27009517的博客
12-18 2632
论文和相关代码 ：https://paperswithcode.com/ Caiming Xiong http://www.stat.ucla.edu/~caiming/ 论文，代码，博客 肖小粤的啵嗑 http://xiaoxiaoyue.me/ 博客记录 小土刀 https://wdxtub.com/ 博客记录 博客 http://ruder.io/multi-task/...
Lecture1-2What is Machine Learning?
zypandora的博客
09-11 525
From Learning to Machine Learning
Lecture5-3Effective number of hypotheses
zypandora的博客
09-29 505
Dichotomies={hypothesis h:→{x,o}}\mathcal{H} = \{\textrm{hypothesis } h: \mathcal{X} \rightarrow \{x,o\}\}Call: h(x1,x2,...,xN)=(h(x1),h(x2),...,h(xN))∈{x,o}N h(\mathbf{x_1, x_2, ... , x_N}) = (h(\m
Lecture3-1Learning with different output space
zypandora的博客
09-20 359
Binary ClassificationMulticlass Classification: Coin Recognition ProblemRegression: Patient Recovery Prediction ProblemStructured Learning: TaggingMini Summary
Lecture2-3Guarantee of PLA
zypandora的博客
09-19 327
Linear Separable If PLA halts(no mistakes),(necessary condition) \mathcal{D} allows some w\mathbf{w} to make no mistake Call \mathcal{D} linear separable Linear Separable ⟺\mathcal{D} \Longleftrigh
Lecture2-2Perceptron Learning Algorithm
zypandora的博客
09-19 326
Select gg from \mathcal{H} =\mathcal{H}=all possible perceptrons, g=?g = ? want g≈fg \approx f almost necessary: g≈fg \approx f on \mathcal{D}, ideally g(xn)=f(xn)=yng(x_n)=f(x_n)=y_n Difficult: \m
Lecture6-3Bounding Function: inductive cases
zypandora的博客
10-12 310
Estimating B(4,3)B(4,3) B(4,3)B(4,3) = 11 after searching all 16 possibilities. Reorgnizing dichotomies of B(4,3)B(4,3)Check for B(3,3)B(3,3)α+β\alpha + \beta: # of different x1,x2,x3x_1, x_2, x_3 co
Lecture6-2Bounding Function
zypandora的博客
10-12 300
Bounding Function B(N,k)B(N,k) Maximum possible mm_{\mathcal{H}} when break point = kCombinatiorial quantity: no shatter of any length k points in N points irrelevant of details of \mathcal{H} or a
Lecture6-1Restriction of Break Point
zypandora的博客
10-12 299
What must be true when min break point is k=2k=2? N=1N=1, every m(N)=2m_{\mathcal{H}}(N)=2 by definition N=1N=1, every m(N)<4m_{\mathcal{H}}(N)<4 by definition N=3N=3?? Must coform that m(N)<4m_{\
Lecture5-2Effective number of lines
zypandora的博客
09-29 288
Where did MM come from?Where did Union Bound fail? Union Bound Over-Estimating!! Can we group hypotheses by kind?How many lines are there? For only one point x1x_1, how many different lines are t
Lecture 1-5ML and Other Fields
zypandora的博客
09-12 269
ML and Data MiningMachine Learning Use data to compute hypothesis gg that approximates target ff Text Mning Use (huge) data to find property that is interestingif ‘interesting property’ same as ‘hy
Lecture 3-1：程序正确性和性能分析1 基本语句确定和函数式编程原理
在Lecture 3-1中，我们学习了关于程序的正确性和性能分析的内容。首先，我们介绍了一个函数f(k)，它满足f(0) = 1和对于任意的k，f(k) = k。这个函数可以用来说明程序的正确性。 接下来，我们学习了如何找出算法中的...
评论
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