Muutke küpsiste eelistusi

Essentials of Pattern Recognition: An Accessible Approach [Kõva köide]

4.00/5 (2 hinnangut Goodreads-ist)
(Nanjing University, China)
  • Formaat: Hardback, 398 pages, kõrgus x laius x paksus: 250x174x24 mm, kaal: 960 g, Worked examples or Exercises
  • Ilmumisaeg: 19-Nov-2020
  • Kirjastus: Cambridge University Press
  • ISBN-10: 1108483461
  • ISBN-13: 9781108483469
Teised raamatud teemal:
Essentials of Pattern Recognition: An Accessible ApproachSuurem pilt
  • Formaat: Hardback, 398 pages, kõrgus x laius x paksus: 250x174x24 mm, kaal: 960 g, Worked examples or Exercises
  • Ilmumisaeg: 19-Nov-2020
  • Kirjastus: Cambridge University Press
  • ISBN-10: 1108483461
  • ISBN-13: 9781108483469
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9781108483469.html
Märksõnad:
This textbook introduces fundamental concepts, major models, and popular applications of pattern recognition for a one-semester undergraduate course. To ensure student understanding, the text focuses on a relatively small number of core concepts with an abundance of illustrations and examples. Concepts are reinforced with hands-on exercises to nurture the student's skill in problem solving. New concepts and algorithms are framed by real-world context and established as part of the big picture introduced in an early chapter. A problem-solving strategy is employed in several chapters to equip students with an approach for new problems in pattern recognition. This text also points out common errors that a new player in pattern recognition may encounter, and fosters the ability for readers to find useful resources and independently solve a new pattern recognition task through various working examples. Students with an undergraduate understanding of mathematical analysis, linear algebra, and probability will be well prepared to master the concepts and mathematical analysis presented here.

Introduces fundamental concepts, major models, and popular applications of pattern recognition for a one-semester undergraduate course. The text focuses on a relatively small number of core concepts with an abundance of illustrations and examples and provides extensive practice through student exercises.

Arvustused

'I highly recommend this book to all those computer-science students who mainly focus on deep learning: this is the book they should read, where they can learn the fundamentals and the big picture of pattern recognition, which will benefit them in the long run.' Jianfei Cai, Monash University 'Dr. Wu has written a valuable book that could not be more timely: the commoditization of machine learning is putting increasingly powerful tools for working with data in the hands of an increasingly broad population of users and practitioners. However, using these tools correctly and interpreting their outputs properly still require significant expertise. This book fills the gap between the classic pattern-recognition texts that assume a substantial amount of background knowledge and preparation and the innumerable internet blog posts which are highly accessible but often superficial. I am sure this self-contained and useful book will enjoy widespread adoption, and I recommend it highly.' James M. Rehg, Georgia Institute of Technology 'This well-designed book is both accessible and substantial in content. I highly recommend it as a textbook as well as for self-study.' Zhi-Hua Zhou, Nanjing University

Muu info

An accessible undergraduate introduction to the concepts and methods in pattern recognition, machine learning and deep learning.
List of Figures
ix
List of Tables
xi
Preface xiii
Notation xvi
Part I Introduction and Overview
1 Introduction
3(12)
1.1 An Example: Autonomous Driving
4(2)
1.2 Pattern Recognition and Machine Learning
6(6)
Exercises
12(3)
2 Mathematical Background
15(29)
2.1 Linear Algebra
15(10)
2.2 Probability
25(9)
2.3 Optimization and Matrix Calculus
34(5)
2.4 Complexity of Algorithms
39(1)
2.5 Miscellaneous Notes and Additional Resources
40(1)
Exercises
41(3)
3 Overview of a Pattern Recognition System
44(19)
3.1 Face Recognition
44(1)
3.2 A Simple Nearest Neighbor Classifier
45(4)
3.3 The Ugly Details
49(3)
3.4 Making Assumptions and Simplifications
52(7)
3.5 A Framework
59(1)
3.6 Miscellaneous Notes and Additional Resources
59(2)
Exercises
61(2)
4 Evaluation
63(38)
4.1 Accuracy and Error in the Simple Case
63(7)
4.2 Minimizing the Cost/Loss
70(3)
4.3 Evaluation in Imbalanced Problems
73(6)
4.4 Can We Reach 100% Accuracy?
79(6)
4.5 Confidence in the Evaluation Results
85(7)
4.6 Miscellaneous Notes and Additional Resources
92(1)
Exercises
93(8)
Part II Domain-Independent Feature Extraction
5 Principal Component Analysis
101(22)
5.1 Motivation
101(3)
5.2 PCA to Zero-Dimensional Subspace
104(2)
5.3 PCA to One-Dimensional Subspace
106(4)
5.4 PCA for More Dimensions
110(1)
5.5 The Complete PCA Algorithm
110(1)
5.6 Variance Analysis
111(4)
5.7 When to Use or Not to Use PCA?
115(3)
5.8 The Whitening Transform
118(1)
5.9 Eigen-Decomposition vs. SVD
118(1)
5.10 Miscellaneous Notes and Additional Resources
119(1)
Exercises
119(4)
6 Fisher's Linear Discriminant
123(20)
6.1 FLD for Binary Classification
125(7)
6.2 FLD for More Classes
132(3)
6.3 Miscellaneous Notes and Additional Resources
135(1)
Exercises
136(7)
Part III Classifiers and Tools
7 Support Vector Machines
143(30)
7.1 The Key SVM Idea
143(4)
7.2 Visualizing and Calculating the Margin
147(3)
7.3 Maximizing the Margin
150(2)
7.4 The Optimization and the Solution
152(5)
7.5 Extensions for Linearly Inseparable and Multiclass Problems
157(4)
7.6 Kernel SVMs
161(6)
7.7 Miscellaneous Notes and Additional Resources
167(1)
Exercises
167(6)
8 Probabilistic Methods
173(23)
8.1 The Probabilistic Way of Thinking
173(2)
8.2 Choices
175(3)
8.3 Parametric Estimation
178(6)
8.4 Nonparametric Estimation
184(7)
8.5 Making Decisions
191(1)
8.6 Miscellaneous Notes and Additional Resources
192(1)
Exercises
192(4)
9 Distance Metrics and Data Transformations
196(23)
9.1 Distance Metrics and Similarity Measures
196(11)
9.2 Data Transformation and Normalization
207(6)
9.3 Miscellaneous Notes and Additional Resources
213(1)
Exercises
213(6)
10 Information Theory and Decision Trees
219(26)
10.1 Prefix Code and Huffman Tree
219(2)
10.2 Basics of Information Theory
221(5)
10.3 Information Theory for Continuous Distributions
226(5)
10.4 Information Theory in ML and PR
231(3)
10.5 Decision Trees
234(5)
10.6 Miscellaneous Notes and Additional Resources
239(1)
Exercises
239(6)
Part IV Handling Diverse Data Formats
11 Sparse and Misaligned Data
245(21)
11.1 Sparse Machine Learning
245(9)
11.2 Dynamic Time Warping
254(8)
11.3 Miscellaneous Notes and Additional Resources
262(1)
Exercises
262(4)
12 Hidden Markov Model
266(27)
12.1 Sequential Data and the Markov Property
266(8)
12.2 Three Basic Problems in HMM Learning
274(1)
12.3 α, β and the Evaluation Problem
275(5)
12.4 γ, δ, &phsi; and the Decoding Problem
280(3)
12.5 ζ and Learning HMM Parameters
283(3)
12.6 Miscellaneous Notes and Additional Resources
286(1)
Exercises
287(6)
Part V Advanced Topics
13 The Normal Distribution
293(23)
13.1 Definition
293(3)
13.2 Notation and Parameterization
296(1)
13.3 Linear Operation and Summation
297(2)
13.4 Geometry and the Mahalanobis Distance
299(1)
13.5 Conditioning
300(2)
13.6 Product of Gaussians
302(1)
13.7 Application I: Parameter Estimation
303(2)
13.8 Application II: Kalman Filter
305(2)
13.9 Useful Math in This
Chapter
307(5)
Exercises
312(4)
14 The Basic Idea behind Expectation-Maximization
316(17)
14.1 GMM: A Worked Example
316(5)
14.2 An Informal Description of the EM Algorithm
321(1)
14.3 The Expectation-Maximization Algorithm
321(7)
14.4 EM for GMM
328(2)
14.5 Miscellaneous Notes and Additional Resources
330(1)
Exercises
331(2)
15 Convolutional Neural Networks
333(32)
15.1 Preliminaries
334(2)
15.2 CNN Overview
336(5)
15.3 Layer Input, Output, and Notation
341(1)
15.4 The ReLU Layer
342(2)
15.5 The Convolution Layer
344(12)
15.6 The Pooling Layer
356(3)
15.7 A Case Study: The VGG16 Net
359(2)
15.8 Hands-On CNN Experiences
361(1)
15.9 Miscellaneous Notes and Additional Resources
362(1)
Exercises
362(3)
Bibliography 365(14)
Index 379
Jianxin Wu is a professor in the Department of Computer Science and Technology and the School of Artificial Intelligence at Nanjing University, China. He received his B.S. and M.S. degrees in computer science from Nanjing University and his Ph.D. degree in computer science from the Georgia Institute of Technology. Professor Wu has served as an area chair for the conference on Computer Vision and Pattern Recognition (CVPR), the International Conference on Computer Vision (ICCV), and the AAAI Conference on Artificial Intelligence, and he is an associate editor for the Pattern Recognition journal. His research interests are computer vision and machine learning.