Table of Contents

Cover

Preface

PART 1: Computational Data Analysis and Methods

1 Semi-supervised Learning Based on Distributionally Robust Optimization
1. 1.1. Introduction
2. 1.2. Alternative semi-supervised learning procedures
3. 1.3. Semi-supervised learning based on DRO
4. 1.4. Error improvement of our SSL-DRO formulation
5. 1.5. Numerical experiments
6. 1.6. Discussion on the size of the uncertainty set
7. 1.7. Conclusion
8. 1.8. Appendix: supplementary material: technical details for theorem 1.1
9. 1.9. References
2 Updating of PageRank in Evolving Treegraphs
1. 2.1. Introduction
2. 2.2. Abbreviations and definitions
3. 2.3. Finding components
4. 2.4. Maintaining the level of cycles
5. 2.5. Calculating PageRank
6. 2.6. PageRank of a tree with at least a cycle after addition of an edge
7. 2.7. Updating PageRank of evolving treegraph with cyclic components
8. 2.8. Aggregation/disaggregation methods of stochastic matrices
9. 2.9. Numerical experiment
10. 2.10. Procedure to compute PageRank
11. 2.11. Conclusion
12. 2.12. Acknowledgements
13. 2.13. References
3 Exploring The Relationship Between Ordinary PageRank, Lazy PageRank and Random Walk with Backstep PageRank for Different Graph Structures
1. 3.1. Introduction
2. 3.2. Notations and basic concepts
3. 3.3. Mathematical relationships between variants of PageRank
4. 3.4. Convergence rates of the variants of PageRank
5. 3.5. Comparison of ranking behaviors for the variants of PageRank
6. 3.6. Conclusion
7. 3.7. Acknowledgements
8. 3.8. References
4 On the Behavior of Alternative Splitting Criteria for CUB Model-based Trees
1. 4.1. Introduction
2. 4.2. Cubremot
3. 4.3. Application and comparison
4. 4.4. Further developments
5. 4.5. References
5 Investigation on Life Satisfaction Through (Stratified) Chain Regression Graph Models
1. 5.1. Introduction
2. 5.2. Methodology
3. 5.3. Application
4. 5.4. Conclusion
5. 5.5. References

PART 2: Classification Data Analysis and Methods

6 Selection of Proximity Measures for a Topological Correspondence Analysis
1. 6.1. Introduction
2. 6.2. Topological correspondence
3. 6.3. Application to real data and empirical results
4. 6.4. Conclusion and perspectives
5. 6.5. Appendix
6. 6.6. References
7 Support Vector Machines: A Review and Applications in Statistical Process Monitoring
1. 7.1. Introduction
2. 7.2. Review of the literature
3. 7.3. Application
4. 7.4. Conclusion
5. 7.5. Acknowledgement
6. 7.6. References
8 Binary Classification Techniques: An Application on Simulated and Real Bio-medical Data
1. 8.1. Introduction
2. 8.2. Related work
3. 8.3. Materials and methods
4. 8.4. Experimental evaluation
5. 8.5. Results
6. 8.6. Discussion
7. 8.7. Conclusion
8. 8.8. Acknowledgements
9. 8.9. References
9 Some Properties of the Multivariate Generalized Hyperbolic Models
1. 9.1. Introduction
2. 9.2. The MGH family of distributions and their limiting forms
3. 9.3. The conditional MGH distribution and its limits
4. 9.4. References
10 On Determining the Value of Online Customer Satisfaction Ratings – A Case-based Appraisal
1. 10.1. Introduction
2. 10.2. Incomplete, inconsistent and contradictory results
3. 10.3. Sample size volatility
4. 10.4. Technical inadequacies of the customer score criterion
5. 10.5. Non-standard weighting of survey responses
6. 10.6. Survey bias
7. 10.7. Conclusion
8. 10.8. References
11 Projection Clustering Unfolding: A New Algorithm for Clustering Individuals or Items in a Preference Matrix
1. 11.1. Introduction
2. 11.2. Preference data
3. 11.3. Projection pursuit
4. 11.4. Projection pursuit clustering
5. 11.5. Clustering preference data
6. 11.6. Conclusion
7. 11.7. References

List of Authors

Index

End User License Agreement

List of Tables

Chapter 1

Table 1.1. Numerical experiments for real data sets

Chapter 3

Table 3.1. Top 10 vertices picked by the variants of PageRank when β = 0.01 and ...
Table 3.2. Top 10 vertices picked by the variants of PageRank when β = 0.01 and ...
Table 3.3. Top 10 vertices picked by the variants of PageRank when β = 0.499 and...
Table 3.4. Top 10 vertices picked by the variants of PageRank when β = 0.499 and...

Chapter 5

Table 5.1. Categories of the covariates of S which lead to great positive values...
Table 5.2. Categories of the covariates of S which lead to great positive values...

Chapter 6

Table 6.1. Some proximity measures
Table 6.2. The 2 × 2 contingency table between modalities z^j and z^k
Table 6.3. Contingency table - Enterprise births in France 2016 (in thousands)
Table 6.4. Topological Index of Independence & Kappa test. For a color version o...
Table 6.5. Assignment of the reference measure. For a color version of this tabl...
Table 6.6. The 2 × 2 contingency table – Reference and Jaccard measures
Table 6.7. Similarities S(V_ui, V_uj) & Kappa coefficient For a color version of...

Chapter 7

Table 7.1. 73 papers from 2002 to 2018, arranged by category

Chapter 8

Table 8.1. 2 × 2 contingency table for binary classification health cases. White...

Chapter 10

Table 10.1. Published Which? results, 2016
Table 10.2. Which?’s survey methodology in overview
Table 10.3. Scale used for category items
Table 10.4. Scales used for Satisfaction and Endorsement questions
Table 10.6. Spearman’s rank correlations
Table 10.7. Which?’s scoring system for Satisfaction and Endorsement responses

Chapter 11

Table 11.1. Confusion Matrix
Table 11.2. Median ranking within each cluster

List of Illustrations

Chapter 1

Figure 1.1. Idealization of the way in which the unlabeled predictive variables ...
Figure 1.2. Pictorial representation of the role that the support constraint pla...

Chapter 2

Figure 2.1. A directed graph and corresponding components. L_i denotes the ith la...
Figure 2.2. Example of a graph and corresponding a component from the SCC partit...
Figure 2.3. {v_s} is a feedback vertex set
Figure 2.4. A comparison of convergence of PageRank vector computed by power met...

Chapter 3

Figure 3.1. Random walk with backstep at vertex v_j: p is a parent and c is a chi...
Figure 3.2. Geometric convergence of the variants of PageRanks. For a color vers...
Figure 3.3. A simple line graph V = {v₁, v₂, v₃, v₄}
Figure 3.4. Comparison of the variants of PageRanks for v1. For a color version ...
Figure 3.5. A simple line graph and a complete graph
Figure 3.6. Comparison of the variants of PageRanks for nodes v5 and v7. For a c...

Chapter 4

Figure 4.1. Observed and fitted CUB distributions of work-related stress at the ...
Figure 4.2. CUBREMOT for perceived stress with application of the log-likelihood...
Figure 4.3. Terminal node distributions for the CUBREMOT with log-likelihood spl...
Figure 4.4. CUBREMOT for perceived stress with application of the dissimilarity ...
Figure 4.5. Terminal node distributions for the CUBREMOT with dissimilarity spli...
Figure 4.6. Tree for perceived stress as implemented by the partymob package
Figure 4.7. Tree for perceived stress as implemented by the Rpartscore package

Chapter 5

Figure 5.1. On the left, a CRGM represents C ⊥ D|AB, B ⊥ C|A and AB ⊥ E. On the ...
Figure 5.2. Mosaic plots concerning the variables Working condition, W, and life...
Figure 5.3. Best fitting SCRGM. LRT statistics = 25.68, df = 16, p-value = 0.06,...

Chapter 6

Figure 6.1. RNG example with seven groups-modalities – associated adjacency matr...
Figure 6.2. Hierarchical tree of the proximity measures. For a color version of ...

Chapter 7

Figure 7.1. Stacked barplot for our four categories and publication years of the...
Figure 7.2. Barplot showing that 96% of the CCPR papers handle cases with trends...
Figure 7.3. Left: 38% of the published works belongs to our first category (CCPR...
Figure 7.4. Left: the type of the data that are used to test the proposed method...
Figure 7.5. Left: The kernel functions used for the SVM models. Right: The algor...
Figure 7.6. The one-against-one method is used 48% of the time, the one-against-...
Figure 7.7. Data preprocessing procedures mentioned in the review. From top to b...
Figure 7.8. Left: The purpose of the SVM model in the proposed methods. PR stand...
Figure 7.9. The D-SVM chart applied on the white wine data set produces 11 false...
Figure 7.10. The proposed D_a-SVM chart is a better approach for the particular p...

Chapter 8

Figure 8.1. Value distribution between positive and negative samples for k = 5. ...
Figure 8.2. Performance of the classification algorithms and indices. For a colo...
Figure 8.3. AUC and sensitivity for different k values. For a color version of t...
Figure 8.4. AUC and sensitivity for a varying number of features. For a color ve...
Figure 8.5. AUC and sensitivity for a varying population size. For a color versi...
Figure 8.6. AUC and sensitivity for a varying positive:negative ratio. For a col...
Figure 8.7. AUC and sensitivity for varying λ values. For a color version of thi...
Figure 8.8. Performance of the classification algorithms and indices on the data...
Figure 8.9. AUC and sensitivity for different discretization levels (k). For a c...

Chapter 10

Figure 10.1. Sample size by hotel chain
Figure 10.2. Laney P’ chart of customer score by hotel chain. For a color versio...
Figure 10.2
Figure 10.3. Factor analysis – selective output. For a color version of this fig...
Figure 10.4. Published Which? customer scores based on (1, 2, 4, 8, 16) scoring
Figure 10.5. Alternative customer scores based on a (1, 2, 3, 4, 5) scoring sche...
Figure 10.6. Volunteer panel web survey protocol (source: Lee and Valliant 2009)

Chapter 11

Figure 11.1. Projection clustering unfolding solution. Items are coded as: toast...
Figure 11.2. Procrustes configuration plot: objects’ points unfolding configurat...
Figure 11.3. Procrustes configuration plot: individuals’ points unfolding config...
Figure 11.4. Procrustes configuration plot: unfolding (green) versus clustered u...

First published 2020 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address:

ISTE Ltd
27-37 St George’s Road
London SW19 4EU
UK

www.iste.co.uk

John Wiley & Sons, Inc.
111 River Street
Hoboken, NJ 07030
USA

www.wiley.com

The rights of Andreas Makrides, Alex Karagrigoriou and Christos H. Skiadas to be identified as the authors of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act 1988.

Library of Congress Control Number: 2019957555

British Library Cataloguing-in-Publication Data

A CIP record for this book is available from the British Library

ISBN 978-1-78630-534-3

Preface

Thanks to the important work of the authors and contributors we have developed this collective volume on “Data Analysis and Applications: Computational, Classification, Financial, Statistical and Stochastic Methods”.

Data analysis as an area of importance has grown exponentially, especially during the past couple of decades. This can be attributed to a rapidly growing computer industry and the wide applicability of computational techniques, in conjunction with new advances of analytic tools. This being the case, the need for literature that addresses this is self-evident. New publications appear as printed or e-books covering the need for information from all fields of science and engineering thanks to the wide applicability of data analysis and statistic packages.

The book is a collective work by a number of leading scientists, analysts, engineers, mathematicians and statisticians who have been working on the front end of data analysis. The chapters included in this collective volume represent a cross-section of current concerns and research interests in the above-mentioned scientific areas. This volume is divided into two parts with a total of 11 chapters in a form to provide the reader with both theoretical and applied information on data analysis methods, models and techniques along with appropriate applications.

Part I focuses on Computational Data Analysis and Methods and includes five chapters on “Semi-supervised Learning Based on Distributionally Robust Optimization” authored by Jose Blanchet and Yang Kang, “Updating of PageRank in Evolving Treegraphs” by Benard Abola, Pitos Seleka Biganda, Christopher Engström, John Magero Mango, Godwin Kakuba and Sergei Silvestrov, “Exploring the Relationship Between Ordinary PageRank, Lazy PageRank and Random Walk with Backstep PageRank for Different Graph Structures” by Pitos Seleka Biganda, Benard Abola, Christopher Engström, John Magero Mango, Godwin Kakuba and Sergei Silvestrov, “On the Behavior of Alternative Splitting Criteria for CUB Model-based Trees” by Carmela Cappelli, Rosaria Simone and Francesca di Iorio and “Investigation on Life Satisfaction Through (Stratified) Chain Regression Graph Models” by Federica Nicolussi and Manuela Cazzaro.

Part II covers the area of Classification Data Analysis and Methods and includes six chapters on “Selection of Proximity Measures for a Topological Correspondence Analysis” by Rafik Abdesselam, “Support Vector Machines: A Review and Applications in Statistical Process Monitoring” by Anastasios Apsemidis and Stelios Psarakis, “Binary Classification Techniques: An Application on Simulated and Real Bio-medical Data” by Fragkiskos G. Bersimis, Iraklis Varlamis, Malvina Vamvakari and Demosthenes B. Panagiotakos, “Some Properties of the Multivariate Generalized Hyperbolic Models” by Stergios B. Fotopoulos, Venkata K. Jandhyala and Alex Paparas, “On Determining the Value of Online Customer Satisfaction Ratings – A Case-based Appraisal” by Jim Freeman and “Projection Clustering Unfolding: A New Algorithm for Clustering Individuals or Items in a Preference Matrix” by Mariangela Sciandra, Antonio D’Ambrosio and Antonella Plaia.

We wish to thank all the authors for their insights and excellent contributions to this book. We would like to acknowledge the assistance of all involved in the reviewing process of the book, without whose support this could not have been successfully completed. Finally, we wish to express our thanks to the secretariat and, of course, the publishers. It was a great pleasure to work with them in bringing to life this collective volume.

Andreas MAKRIDES

Rouen, France

Alex KARAGRIGORIOU

Samos, Greece

Christos H. SKIADAS

Athens, Greece

January 2020

Data Analysis and Applications 3

Computational, Classification, Financial, Statistical and Stochastic Methods

Preface

PART 1
Computational Data Analysis and Methods