Machine Learning: Theory and Applications
- 1st Edition, Volume 31 - May 16, 2013
- Imprint: North Holland
- Editors: C.R. Rao, Venu Govindaraju
- Language: English
- Paperback ISBN:9 7 8 - 1 - 4 9 3 3 - 0 2 4 3 - 7
- Hardback ISBN:9 7 8 - 0 - 4 4 4 - 5 3 8 5 9 - 8
- eBook ISBN:9 7 8 - 0 - 4 4 4 - 5 3 8 6 6 - 6
Statistical learning and analysis techniques have become extremely important today, given the tremendous growth in the size of heterogeneous data collections and the ability to… Read more
Purchase options
Institutional subscription on ScienceDirect
Request a sales quoteStatistical learning and analysis techniques have become extremely important today, given the tremendous growth in the size of heterogeneous data collections and the ability to process it even from physically distant locations. Recent advances made in the field of machine learning provide a strong framework for robust learning from the diverse corpora and continue to impact a variety of research problems across multiple scientific disciplines. The aim of this handbook is to familiarize beginners as well as experts with some of the recent techniques in this field.The Handbook is divided in two sections: Theory and Applications, covering machine learning, data analytics, biometrics, document recognition and security.
- Very relevant to current research challenges faced in various fields
- Self-contained reference to machine learning
- Emphasis on applications-oriented techniques
Computer scientists, artificial intelligence researchers, biometricians, and statisticians
Contributors: Vol. 31
Preface to Handbook Volume – 31
Introduction
1 Part I—Theoretical aspects
2 Part II—Object recognition
3 Part III—Biometric systems
4 Part IV—Document analysis
Part I: Theoretical Analysis
Chapter 1. The Sequential Bootstrap
1 Introduction
2 A sequential bootstrap resampling scheme
3 Bootstrapping empirical measures with a random sample size
4 Convergence rates for the sequential bootstrap
5 Second-order correctness of the sequential bootstrap
6 Concluding remarks
Acknowledgments
References
Chapter 2. The Cross-Entropy Method for Estimation
1 Introduction
2 Estimation setting
3 Extensions
Acknowledgement
References
Chapter 3. The Cross-Entropy Method for Optimization
1 Introduction
2 From estimation to optimization
3 Applications to combinatorial optimization
4 Continuous optimization
5 Summary
References
Chapter 4. Probability Collectives in Optimization
1 Introduction
2 Delayed sampling theory
3 Delayed sampling experiments
4 Immediate sampling theory
5 Immediate sampling experiments
6 Conclusion
References
Chapter 5. Bagging, Boosting, and Random Forests Using R
1 Introduction
2 Data sets and rationale
3 Bagging
4 Boosting
5 Do Bagging and Boosting really work?
6 What is a classification tree?
7 Classification tree versus logistic regression
8 Random forest
9 Random forest, genetics, and cross-validation
10 Regression trees
11 Boosting using the R package, ada
12 Epilog
References
Chapter 6. Matching Score Fusion Methods
1 Introduction
2 Matching systems
3 Selected approaches to fusion in matching systems
4 Operating modes of matching systems
5 Complexity types of classifier combination methods
6 Modeling matching score dependencies
7 Score combination applications
8 Conclusion
References
Part II: Object Recognition
Chapter 7. Statistical Methods on Special Manifolds for Image and Video Understanding
1 Introduction
2 Some motivating examples
3 Differential geometric tools
4 Common manifolds arising in image analysis
5 Applications in image analysis
6 Summary and discussion
Acknowledgments
References
Chapter 8. Dictionary-Based Methods for Object Recognition∗
1 Introduction
2 Sparse representation
3 Dictionary learning
4 Concluding remarks
References
Chapter 9. Conditional Random Fields for Scene Labeling
1 Introduction
2 Overview of CRF
3 Scene parsing
4 More recent implementations of CRF scene labelings
5 Conclusion and future directions
References
Chapter 10. Shape-Based Image Classification and Retrieval
1 Introduction
2 Prior work
3 Classification and retrieval models
4 Features
5 Classification experiments
6 Retrieval
7 Multiple class labels
8 Summary and conclusions
References
Chapter 11. Visual Search: A Large-Scale Perspective
1 Introduction
2 When is big data important?
3 Information extraction and representation
4 Matching images
5 Practical considerations: memory footprint and speed
6 Benchmark data sets
7 Closing remarks
References
Part III: Biometric Systems
Chapter 12. Video Activity Recognition by Luminance Differential Trajectory and Aligned Projection Distance
1 Introduction
2 Related work
3 Problem formulation
4 DLFT and LAPD solutions
5 Experiments
6 Conclusion
References
Chapter 13. Soft Biometrics for Surveillance: An Overview
1 Introduction
2 Performance metrics
3 Incorporating soft biometrics in a fusion framework
4 Human identification using soft biometrics
5 Predicting gender from face images
6 Applications
7 Conclusion
References
Chapter 14. A User Behavior Monitoring and Profiling Scheme for Masquerade Detection
1 Introduction
2 Related work
3 Support Vector Machines (SVMs)
4 Data collection, feature extraction, and feature vector generation
5 Experimental design
6 Discussion and conclusion
Acknowledgments
References
Chapter 15. Application of Bayesian Graphical Models to Iris Recognition
1 Introduction
2 Gabor wavelet-based matching
3 Correlation filter-based iris matching
4 Bayesian graphical model for iris recognition
5 Summary
Acknowledgments
References
Part IV: Document Analysis
Chapter 16. Learning Algorithms for Document Layout Analysis
1 Introduction
2 Pixel classification
3 Zone classification
4 Connected component classification
5 Text region segmentation
6 Region classification
7 Functional labeling
8 Conclusion
References
Chapter 17. Hidden Markov Models for Off-Line Cursive Handwriting Recognition
1 Introduction
2 Serialization of handwriting images
3 HMM-based text line recognition
4 Outlook and conclusions
Acknowledgment
References
Chapter 18. Machine Learning in Handwritten Arabic Text Recognition
1 Introduction
2 Arabic script—challenges for recognition
3 Learning paradigms
4 Features for text recognition
5 Models for recognition
6 Conclusion
References
Chapter 19. Manifold Learning for the Shape-Based Recognition of Historical Arabic Documents
1 Introduction
2 Problem statement
3 Manifold learning
4 Feature extraction
5 Experimental results
6 Conclusion and future prospects
Acknowledgments
References
Chapter 20. Query Suggestion with Large Scale Data
1 Introduction
2 Terminology
3 Approaches to generation of Query Suggestions
4 Evaluation methods of QS
5 Properties of large scale data
6 Query Suggestion in practice
7 Closing remarks
References
Subject Index
- Edition: 1
- Volume: 31
- Published: May 16, 2013
- No. of pages (Hardback): 552
- Imprint: North Holland
- Language: English
- Paperback ISBN: 9781493302437
- Hardback ISBN: 9780444538598
- eBook ISBN: 9780444538666
CR
C.R. Rao
book “Ancient Inhabitants of Jebel Moya” published by the Cambridge Press under the joint authorship of Rao and two anthropologists. On the basis of work done at CU during the two year period, 1946-1948, Rao earned a Ph.D. degree and a few years later Sc.D. degree of CU and the rare honor of life fellowship of Kings College, Cambridge.
He retired from ISI in 1980 at the mandatory age of 60 after working for 40 years during which period he developed ISI as an international center for statistical education and research. He also took an active part in establishing state statistical bureaus to collect local statistics and transmitting them to Central Statistical Organization in New Delhi. Rao played a pivitol role in launching undergraduate and postgraduate courses at ISI. He is the author of 475 research publications and several breakthrough papers contributing to statistical theory and methodology for applications to problems in all areas of human endeavor. There are a number of classical statistical terms named after him, the most popular of which are Cramer-Rao inequality, Rao-Blackwellization, Rao’s Orthogonal arrays used in quality control, Rao’s score test, Rao’s Quadratic Entropy used in ecological work, Rao’s metric and distance which are incorporated in most statistical books.
He is the author of 10 books, of which two important books are, Linear Statistical Inference which is translated into German, Russian, Czec, Polish and Japanese languages,and Statistics and Truth which is translated into, French, German, Japanese, Mainland Chinese, Taiwan Chinese, Turkish and Korean languages.
He directed the research work of 50 students for the Ph.D. degrees who in turn produced 500 Ph.D.’s. Rao received 38 hon. Doctorate degree from universities in 19 countries spanning 6 continents. He received the highest awards in statistics in USA,UK and India: National Medal of Science awarded by the president of USA, Indian National Medal of Science awarded by the Prime Minister of India and the Guy Medal in Gold awarded by the Royal Statistical Society, UK. Rao was a recipient of the first batch of Bhatnagar awards in 1959 for mathematical sciences and and numerous medals in India and abroad from Science Academies. He is a Fellow of Royal Society (FRS),UK, and member of National Academy of Sciences, USA, Lithuania and Europe. In his honor a research Institute named as CRRAO ADVANCED INSTITUTE OF MATHEMATICS, STATISTICS AND COMPUTER SCIENCE was established in the campus of Hyderabad University.
Affiliations and expertise
University of Hyderabad Campus, IndiaVG
Venu Govindaraju
Dr. Venu Govindaraju, SUNY Distinguished Professor of Computer Science and Engineering, is the Vice President of Research and Economic Development of the University at Buffalo and founding director of the Center for Unified Biometrics and Sensors. He received his Bachelor’s degree with honors from the Indian Institute of Technology (IIT) in 1986, and his Ph.D. from UB in 1992. His research focus is on machine learning and pattern recognition in the domains of Document Image Analysis and Biometrics. Dr. Govindaraju has co-authored about 400 refereed scientific papers. His seminal work in handwriting recognition was at the core of the first handwritten address interpretation system used by the US Postal Service. He was also the prime technical lead responsible for technology transfer to the Postal Services in US, Australia, and UK. He has been a Principal or Co-Investigator of sponsored projects funded for about 65 million dollars. Dr. Govindaraju has supervised the dissertations of 30 doctoral students. He has served on the editorial boards of premier journals such as the IEEE Transactions on Pattern Analysis and Machine Intelligence and is currently the Editor-in-Chief of the IEEE Biometrics Council Compendium. Dr. Govindaraju is a Fellow of the ACM (Association of Computing Machinery), IEEE (Institute of Electrical and Electronics Engineers), AAAS (American Association for the Advancement of Science), the IAPR (International Association of Pattern Recognition), and the SPIE (International Society of Optics and Photonics). He is recipient of the 2004 MIT Global Indus Technovator award and the 2010 IEEE Technical Achievement award.
Affiliations and expertise
The State University of New York, Buffalo, NY, USARead Machine Learning: Theory and Applications on ScienceDirect