版权说明:本文档由用户提供并上传,收益归属内容提供方,若内容存在侵权,请进行举报或认领
文档简介
1、Prof. Liqing ZhangDept. Computer Science & Engineering, Shanghai Jiaotong UniversityStatistical Learning & InferenceBooks and ReferencesTrevor Hastie Robert Tibshirani Jerome Friedman , The Elements of statistical Learning: Data Mining, Inference, and Prediction, 2001, Springer-VerlagVladimir N. Vap
2、nik, The Nature of Statistical Learning Theory, 2nd ed., Springer, 2000S. Mendelson, A few notes on Statistical Learning Theory in Advanced Lectures in Machine Learning: Machine Learning Summer School 2002, S. Mendelson and A. J. Smola (eds), Lecture Notes in Computer Science, 2600, Springer, 2003M.
3、 Vidyasagar, Learning and generalization: with applications to neural networks, 2nd ed., Springer, 2003Overview of the CourseIntroductionOverview of Supervised LearningLinear Method for Regression and ClassificationBasis Expansions and RegularizationKernel MethodsModel Selections and InferenceSuppor
4、t Vector MachineBayesian InferenceUnsupervised LearningWhy Statistical Learning?我门被信息淹没,但却缺乏知识。- R. Roger恬静的统计学家改动了我们的世界;不是经过发现新的现实或者开发新技术,而是经过改动我们的推理、实验和观念的构成方式。- I. Hacking问题:为什么如今的计算机处置智能信息效率很低?图像、视频、音频认知言语ML: SARS Risk PredictionSARS RiskAgeGenderBlood PressureChest X-RayPre-Hospital AttributesA
5、lbuminBlood pO2White CountRBC CountIn-Hospital AttributesML: Auto Vehicle NavigationSteering DirectionProtein FoldingThe Scale of Biomedical Data计算科学与脑科学计算机信息处置基于逻辑的计算和数据分别数据处置与存储简单智能信息处置复杂、慢认知才干弱信息处置方式:逻辑概念统计信息大脑信息处置基于统计信息的计算计算和数据集成一体数据处置与存储未知智能信息处置简单、快速认知才干强信息处置方式:统计信息概念逻辑Function Estimation Model
6、The Function Estimation Model of learning examples:Generator (G) generates observations x (typically in Rn), independently drawn from some fixed distribution F(x)Supervisor (S) labels each input x with an output value y according to some fixed distribution F(y|x)Learning Machine (LM) “learns from an
7、 i.i.d. l-sample of (x,y)-pairs output from G and S, by choosing a function that best approximates S from a parameterised function class f(x,), where is in the parameter setFunction Estimation ModelKey concepts: F(x,y), an i.i.d. k-sample on F, functions f(x,) and the equivalent representation of ea
8、ch f using its index xGSLMyyThe loss functional (L, Q)the error of a given function on a given exampleThe risk functional (R)the expected loss of a given function on an example drawn from F(x,y) the (usual concept of) generalisation error of a given function The Problem of Risk Minimization The Prob
9、lem of Risk MinimizationThree Main Learning ProblemsPattern Recognition:Regression Estimation:Density Estimation:General FormulationThe Goal of LearningGiven an i.i.d. k-sample z1, zk drawn from a fixed distribution F(z)For a function class loss functionals Q (z ,), with in We wish to minimise the r
10、isk, finding a function *General FormulationThe Empirical Risk Minimization (ERM) Inductive PrincipleDefine the empirical risk (sample/training error):Define the empirical risk minimiser:ERM approximates Q (z ,*) with Q (z ,k) the Remp minimiserthat is ERM approximates * with kLeast-squares and Maxi
11、mum-likelihood are realisations of ERM4 Issues of Learning TheoryTheory of consistency of learning processesWhat are (necessary and sufficient) conditions for consistency (convergence of Remp to R) of a learning process based on the ERM Principle?Non-asymptotic theory of the rate of convergence of l
12、earning processesHow fast is the rate of convergence of a learning process?Generalization ability of learning processesHow can one control the rate of convergence (the generalization ability) of a learning process?Constructing learning algorithms (i.e. the SVM)How can one construct algorithms that c
13、an control the generalization ability?Change in Scientific MethodologyTRADITIONALFormulate hypothesisDesign experimentCollect dataAnalyze resultsReview hypothesisRepeat/PublishNEWDesign large experimentsCollect large dataPut data in large databaseFormulate hypothesisEvaluate hypothesis on databaseRu
14、n limited experiments Review hypothesisRepeat/PublishLearning & AdaptationIn the broadest sense, any method that incorporates information from training samples in the design of a classifier employs learning.Due to complexity of classification problems, we cannot guess the best classification decisio
15、n ahead of time, we need to learn it.Creating classifiers then involves positing some general form of model, or form of the classifier, and using examples to learn the complete classifier.Supervised learningIn supervised learning, a teacher provides a category label for each pattern in a training se
16、t. These are then used to train a classifier which can thereafter solve similar classification problems by itself.Unsupervised learningIn unsupervised learning, or clustering, there is no explicit teacher or training data. The system forms natural clusters of input patterns and classifiers them base
17、d on clusters they belong to .Reinforcement learningIn reinforcement learning, a teacher only says to classifier whether it is right when suggesting a category for a pattern. The teacher does not tell what the correct category is.ClassificationThe task of the classifier component is to use the featu
18、re vector provided by the feature extractor to assign the object to a category.Classification is the main topic of this course.The abstraction provided by the feature vector representation of the input data enables the development of a largely domain-independent theory of classification.Essentially the classifier divides the feature space into regions corresponding to different categories.ClassificationThe degree of difficulty of the classification pr
温馨提示
- 1. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
- 2. 本站的文档不包含任何第三方提供的附件图纸等,如果需要附件,请联系上传者。文件的所有权益归上传用户所有。
- 3. 本站RAR压缩包中若带图纸,网页内容里面会有图纸预览,若没有图纸预览就没有图纸。
- 4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
- 5. 人人文库网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对用户上传分享的文档内容本身不做任何修改或编辑,并不能对任何下载内容负责。
- 6. 下载文件中如有侵权或不适当内容,请与我们联系,我们立即纠正。
- 7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。
最新文档
- 2026年工艺流程柔性安排方案设计
- 2026年中班组户外活动计划上学期
- 2026年虚拟现实项目开发合同三篇
- 乐山市沙湾区米房沟防洪治理工程水土保持报告表
- 赤峰东山铝业项目(间隔扩建)220千伏送出工程水土保持方案报告表
- 2025-2026学年教学活动方案设计
- 12 美丽的星空 教学设计科学六年级下册冀人版
- 2026年湖北省高考思想政治试卷
- 2.2基本不等式(2)教学设计-高一上学期数学人教A版(2019)必修第一册
- 2024八年级数学下册 第19章 平面直角坐标系19.2平面直角坐标系 1平面直角坐标系教案(新版)冀教版
- 2026河南郑州市郑盐盐业集团有限公司社会招聘7人笔试参考题库及答案详解
- 2026年辽宁锦州海通实业有限公司计划招录28人备考题库及参考答案详解一套
- 2026年黑龙江、吉林、辽宁、内蒙古高考生物试卷(含答案及解析)
- 慢性肾脏病合并心脏病的管理
- 2026统编版小学三年级道德与法治下册期末复习综合测试卷及答案(共三套)
- 海绵城市建设试点项目水土保持方案
- 2026年广东省惠州市初二学业水平地理生物会考试题题库(答案+解析)
- 2026中共深圳市龙岗区委政法委员会招聘聘员4人备考题库(广东)附答案详解ab卷
- 2026年贵州铜仁市初二地生会考真题试卷+解析及答案
- 学校运动猝死案例研究报告
- 重庆公务员 2026真题及答案
评论
0/150
提交评论