没有合适的资源?快使用搜索试试~ 我知道了~
Assessing and Improving Prediction and Classification
需积分: 10 18 下载量 154 浏览量
2017-12-27
13:17:54
上传
评论
收藏 5.36MB PDF 举报
温馨提示
Title: Assessing and Improving Prediction and Classification: Theory and Algorithms in C++ Author: Timothy Masters Length: 517 pages Edition: 1st ed. Language: English Publisher: Apress Publication Date: 2017-12-20 ISBN-10: 1484233352 ISBN-13: 9781484233351
资源推荐
资源详情
资源评论
Assessing and
Improving Prediction
and Classification
Theory and Algorithms in C++
TimothyMasters
Assessing and Improving Prediction and Classication: eory and Algorithms
in C++
ISBN-13 (pbk): 978-1-4842-3335-1 ISBN-13 (electronic): 978-1-4842-3336-8
https://doi.org/10.1007/978-1-4842-3336-8
Library of Congress Control Number: 2017962869
Copyright © 2018 by Timothy Masters
is work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microlms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
Trademarked names, logos, and images may appear in this book. Rather than use a trademark symbol with
every occurrence of a trademarked name, logo, or image we use the names, logos, and images only in an
editorial fashion and to the benet of the trademark owner, with no intention of infringement of the
trademark.
e use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not
identied as such, is not to be taken as an expression of opinion as to whether or not they are subject to
proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of publication,
neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or
omissions that may be made. e publisher makes no warranty, express or implied, with respect to the
material contained herein.
Cover image by Freepik (www.freepik.com)
Managing Director: Welmoed Spahr
Editorial Director: Todd Green
Acquisitions Editor: Steve Anglin
Development Editor: Matthew Moodie
Technical Reviewers: Massimo Nardone and Matt Wiley
Coordinating Editor: Mark Powers
Copy Editor: Kim Wimpsett
Distributed to the book trade worldwide by Springer Science+Business Media NewYork, 233 Spring Street,
6th Floor, NewYork, NY 10013. Phone 1-800-SPRINGER, fax (201) 348-4505, e-mail orders-ny@springer-
sbm.com, or visit www.springeronline.com. Apress Media, LLC is a California LLC and the sole member
(owner) is Springer Science + Business Media Finance Inc (SSBM Finance Inc). SSBM Finance Inc is a
Delaware corporation.
For information on translations, please e-mail rights@apress.com, or visit www.apress.com/
rights-permissions.
Apress titles may be purchased in bulk for academic, corporate, or promotional use. eBook versions and
licenses are also available for most titles. For more information, reference our Print and eBook Bulk Sales
web page at www.apress.com/bulk-sales.
Any source code or other supplementary material referenced by the author in this book is available to
readers on GitHub via the book’s product page, located at www.apress.com/9781484233351. For more
detailed information, please visit www.apress.com/source-code.
Printed on acid-free paper
TimothyMasters
Ithaca, New York, USA
is book is dedicated to Master Hidy Ochiai with the utmost respect,
admiration, and gratitude. His incomparable teaching of Washin-Ryu
karate has raised my condence, my physical ability, and my mental
acuity far beyond anything I could have imagined.
For this I will ever be grateful.
v
About the Author ��������������������������������������������������������������������������������������������������� xiii
About the Technical Reviewers �������������������������������������������������������������������������������xv
Preface ������������������������������������������������������������������������������������������������������������������xvii
Table of Contents
Chapter 1: Assessment ofNumeric Predictions ������������������������������������������������������� 1
Notation......................................................................................................................................... 2
Overview ofPerformance Measures ............................................................................................. 3
Consistency andEvolutionary Stability .................................................................................... 6
Selection Bias andtheNeed forThree Datasets........................................................................... 9
Cross Validation andWalk-Forward Testing ................................................................................ 14
Bias inCross Validation ......................................................................................................... 15
Overlap Considerations.......................................................................................................... 15
Assessing Nonstationarity Using Walk-Forward Testing........................................................ 17
Nested Cross Validation Revisited ......................................................................................... 18
Common Performance Measures ................................................................................................ 20
Mean Squared Error .............................................................................................................. 20
Mean Absolute Error .............................................................................................................. 23
R-Squared ............................................................................................................................. 23
RMS Error .............................................................................................................................. 24
Nonparametric Correlation .................................................................................................... 24
Success Ratios ...................................................................................................................... 26
Alternatives toCommon Performance Measures .................................................................. 27
Stratication forConsistency ...................................................................................................... 27
Condence Intervals ................................................................................................................... 29
The Condence Set ............................................................................................................... 30
Serial Correlation ................................................................................................................... 32
vi
Multiplicative Data ................................................................................................................. 32
Normally Distributed Errors ................................................................................................... 33
Empirical Quantiles asCondence Intervals ............................................................................... 35
Condence Bounds forQuantiles .......................................................................................... 37
Tolerance Intervals ................................................................................................................ 40
Chapter 2: Assessment ofClass Predictions ���������������������������������������������������������� 45
The Confusion Matrix .................................................................................................................. 46
Expected Gain/Loss ............................................................................................................... 46
ROC (Receiver Operating Characteristic) Curves ........................................................................ 48
Hits, False Alarms, andRelated Measures ............................................................................ 48
Computing theROC Curve ..................................................................................................... 50
Area Under theROC Curve ..................................................................................................... 56
Cost andtheROC Curve ........................................................................................................ 59
Optimizing ROC-Based Statistics ................................................................................................ 60
Optimizing theThreshold: Now or Later? .............................................................................. 61
Maximizing Precision ............................................................................................................ 64
Generalized Targets ............................................................................................................... 65
Maximizing Total Gain ............................................................................................................ 66
Maximizing Mean Gain .......................................................................................................... 67
Maximizing theStandardized Mean Gain .............................................................................. 67
Condence inClassication Decisions ....................................................................................... 69
Hypothesis Testing ................................................................................................................. 70
Condence intheCondence ................................................................................................ 75
Bayesian Methods ................................................................................................................. 81
Multiple Classes .................................................................................................................... 85
Hypothesis Testing vs. Bayes’ Method .................................................................................. 86
Final Thoughts onHypothesis Testing ................................................................................... 91
Condence Intervals forFuture Performance ............................................................................. 98
Table of ConTenTs
剩余528页未读,继续阅读
资源评论
DWcsdnNET
- 粉丝: 414
- 资源: 651
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
- ap5030dn-openwrt-ath79-generic-huawei-ap5030dn-initramfs-kernel
- MinIO是一款高性能高可用的文件系统服务,可以用来替换FastDFS minio Docker镜像-v2024.6.29
- Annotations_Train_abstract_v002.zip
- sonatype-nexus3 Docker镜像-v3.9.0
- Java实现基于轻量型卷积神经网络的病虫害分析系统(源码+文档)
- Java毕业设计-基于Springboot轻量型卷积神经网络的病虫害分析系统(源码+文档)
- CIASI 2023测试打分表
- Java毕业设计-基于Springboot植物病虫害分析系统(源码+文档)
- Java毕业设计-基于Springboot的农作物病虫害分析系统(源码+文档)
- CSP竞赛编程基础教程:从入门到精通
- Hacknet.zip
- FPGA开发入门与实践基础教程
- 示波器使用与实验操作基础教程
- JAVA日期转换工具类
- 软考中级基础教程:掌握计算机技术与软件应用
- java下excel导出工具类,支持多个sheet,根据入参配置到处调用即可
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功