Soft Methods for Data Science 1st Edition Maria

Brigida Ferraro pdf download

https://textbookfull.com/product/soft-methods-for-data-
science-1st-edition-maria-brigida-ferraro/

Download more ebook from https://textbookfull.com

We believe these products will be a great fit for you. Click
the link to download now, or visit textbookfull.com
to discover even more!

Java for Data Science 1st Edition Reese

https://textbookfull.com/product/java-for-data-science-1st-
edition-reese/

R for Data Science 1st Edition Garrett Grolemund

https://textbookfull.com/product/r-for-data-science-1st-edition-
garrett-grolemund/

Algorithms for Data Science 1st Edition Brian Steele

https://textbookfull.com/product/algorithms-for-data-science-1st-
edition-brian-steele/

Data Science for Mathematicians 1st Edition Nathan

Carter

https://textbookfull.com/product/data-science-for-
mathematicians-1st-edition-nathan-carter/
Mathematical Problems in Data Science Theoretical and
Practical Methods 1st Edition Li M. Chen

https://textbookfull.com/product/mathematical-problems-in-data-
science-theoretical-and-practical-methods-1st-edition-li-m-chen/

R Programming for Data Science 1st Edition Roger Peng

https://textbookfull.com/product/r-programming-for-data-
science-1st-edition-roger-peng/

Programming Skills for Data Science 1st Edition Michael

Freeman

https://textbookfull.com/product/programming-skills-for-data-
science-1st-edition-michael-freeman/

Cultural Anthropology An Applied Perspective Gary

Ferraro

https://textbookfull.com/product/cultural-anthropology-an-
applied-perspective-gary-ferraro/

Data Science and Social Research II: Methods,

Technologies and Applications Paolo Mariani

https://textbookfull.com/product/data-science-and-social-
research-ii-methods-technologies-and-applications-paolo-mariani/
Advances in Intelligent Systems and Computing 456
Maria Brigida Ferraro
Paolo Giordani
Barbara Vantaggi
Marek Gagolewski
María Ángeles Gil
Przemysław Grzegorzewski
Olgierd Hryniewicz Editors

Soft Methods
for Data
Science
Advances in Intelligent Systems and Computing

Volume 456

Series editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
e-mail: kacprzyk@ibspan.waw.pl
About this Series
The series “Advances in Intelligent Systems and Computing” contains publications on theory,
applications, and design methods of Intelligent Systems and Intelligent Computing. Virtually
all disciplines such as engineering, natural sciences, computer and information science, ICT,
economics, business, e-commerce, environment, healthcare, life science are covered. The list
of topics spans all the areas of modern intelligent systems and computing.
The publications within “Advances in Intelligent Systems and Computing” are primarily
textbooks and proceedings of important conferences, symposia and congresses. They cover
significant recent developments in the field, both of a foundational and applicable character.
An important characteristic feature of the series is the short publication time and world-wide
distribution. This permits a rapid and broad dissemination of research results.

Advisory Board
Chairman
Nikhil R. Pal, Indian Statistical Institute, Kolkata, India
e-mail: nikhil@isical.ac.in
Members
Rafael Bello, Universidad Central “Marta Abreu” de Las Villas, Santa Clara, Cuba
e-mail: rbellop@uclv.edu.cu
Emilio S. Corchado, University of Salamanca, Salamanca, Spain
e-mail: escorchado@usal.es
Hani Hagras, University of Essex, Colchester, UK
e-mail: hani@essex.ac.uk
László T. Kóczy, Széchenyi István University, Győr, Hungary
e-mail: koczy@sze.hu
Vladik Kreinovich, University of Texas at El Paso, El Paso, USA
e-mail: vladik@utep.edu
Chin-Teng Lin, National Chiao Tung University, Hsinchu, Taiwan
e-mail: ctlin@mail.nctu.edu.tw
Jie Lu, University of Technology, Sydney, Australia
e-mail: Jie.Lu@uts.edu.au
Patricia Melin, Tijuana Institute of Technology, Tijuana, Mexico
e-mail: epmelin@hafsamx.org
Nadia Nedjah, State University of Rio de Janeiro, Rio de Janeiro, Brazil
e-mail: nadia@eng.uerj.br
Ngoc Thanh Nguyen, Wroclaw University of Technology, Wroclaw, Poland
e-mail: Ngoc-Thanh.Nguyen@pwr.edu.pl
Jun Wang, The Chinese University of Hong Kong, Shatin, Hong Kong
e-mail: jwang@mae.cuhk.edu.hk

More information about this series at http://www.springer.com/series/11156

Maria Brigida Ferraro Paolo Giordani
•

Barbara Vantaggi Marek Gagolewski

•

María Ángeles Gil Przemysław Grzegorzewski

•

Olgierd Hryniewicz
Editors

Soft Methods for Data

Science

123
Editors
Maria Brigida Ferraro María Ángeles Gil
Department of Statistical Sciences Department of Statistics and Operational
Sapienza University of Rome Research and Mathematics Didactics
Rome University of Oviedo
Italy Oviedo
Spain
Paolo Giordani
Department of Statistical Sciences Przemysław Grzegorzewski
Sapienza University of Rome Department of Stochastic Methods, Systems
Rome Research Institute
Italy Polish Academy of Sciences
Warsaw
Barbara Vantaggi Poland
Department of Basic and Applied Sciences
for Engineering Olgierd Hryniewicz
Sapienza University of Rome Department of Stochastic Methods, Systems
Rome Research Institute
Italy Polish Academy of Sciences
Warsaw
Marek Gagolewski Poland
Department of Stochastic Methods, Systems
Research Institute
Polish Academy of Sciences
Warsaw
Poland

ISSN 2194-5357 ISSN 2194-5365 (electronic)

Advances in Intelligent Systems and Computing
ISBN 978-3-319-42971-7 ISBN 978-3-319-42972-4 (eBook)
DOI 10.1007/978-3-319-42972-4

Library of Congress Control Number: 2016945947

© Springer International Publishing Switzerland 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms 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.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
Preface

This volume is a collection of peer-reviewed papers presented at the 8th

International Conference on Soft Methods in Probability and Statistics—SMPS
2016, held in Rome (Italy) during September 12–14, 2016. The series of biannual
international conferences on Soft Methods in Probability and Statistics (SMPS)
started in Warsaw in 2002. Subsequent events in this series took place in Oviedo
(2004), Bristol (2006), Toulouse (2008), Oviedo/Mieres (2010), Konstanz (2012),
and Warsaw (2014). SMPS 2016 was organized by the Department of Basic and
Applied Sciences for Engineering and the Department of Statistical Sciences,
Sapienza University of Rome, Italy.
Over the last 50 years in different areas such as decision theory, information
processing, and data mining, the interest to extend probability theory and statistics
has grown. The common feature of those attempts is to widen frameworks for
representing different kinds of uncertainty: randomness, imprecision, vagueness,
and ignorance. The scope is to develop more flexible methods to analyze data and
extract knowledge from them. The extension of classical methods consists in
“softening” them by means of new approaches involving fuzzy set theory, possi-
bility theory, rough sets, or having their origin in probability theory itself, like
imprecise probabilities, belief functions, and fuzzy random variables.
Data science aims at developing automated methods to analyze massive amounts
of data and extract knowledge from them. In the recent years the production of data
is dramatically increasing. Every day a huge amount of data coming from every-
where is collected: mobile sensors, sophisticated instruments, transactions, Web
logs, and so forth. This trend is expected to accelerate in the near future. Data
science employs various programming techniques and methods of data wrangling,
data visualization, machine learning, and probability and statistics. The soft
methods proposed in this volume represent a suit of tools in these fields that can
also be useful for data science.
The volume contains 65 selected contributions devoted to the foundation of
uncertainty theories such as probability, imprecise probability, possibility theory,
soft methods for probability and statistics. Some of them are focused on robustness,

v
vi Preface

non-precise data, dependence models with fuzzy sets, clustering, mathematical

models for decision theory and finance.
We would like to thank all contributing authors, organizers of special sessions,
program committee members, and additional referees who made it possible to put
together the attractive program of the conference. We are very grateful to the
plenary speakers, Ana Colubi (University of Oviedo, Spain), Thierry Denoeux
(University of Technology of Compiégne, France), Massimo Marinacci (Bocconi
University, Italy) for their very interesting talks: “On some functional characteri-
zations of (fuzzy) set-valued random elements”, “Beyond Fuzzy, Possibilistic and
Rough: An Investigation of Belief Functions in Clustering” and “A non Bayesian
Approach to Measurement Problems”, respectively. We would like to express our
gratitude also to INDAM-GNAMPA for the financial support. Furthermore, we
would like to thank the editor of the Springer series Advances in Soft Computing,
Prof. Janusz Kacprzyk, and Springer-Verlag for the dedication to the production of
this volume.

Rome Maria Brigida Ferraro

June 2016 Paolo Giordani
Barbara Vantaggi
Marek Gagolewski
María Ángeles Gil
Przemysław Grzegorzewski
Olgierd Hryniewicz
Organization

General Chairs
Maria Brigida Ferraro, Rome, Italy
Paolo Giordani, Rome, Italy
Barbara Vantaggi, Rome, Italy

Executive Board
María Ángeles Gil, Oviedo, Spain
Przemysław Grzegorzewski, Warsaw, Poland
Olgierd Hryniewicz, Warsaw, Poland

Program Committee
Bernard de Baets, Ghent, Belgium
Christian Borgelt, Mieres, Spain
Giulianella Coletti, Perugia, Italy
Ana Colubi, Oviedo, Spain
Ines Couso, Oviedo, Spain
Gert De Cooman, Ghent, Belgium
Thierry Denoeux, Compiégne, France
Didier Dubois, Toulouse, France
Fabrizio Durante, Bolzano, Italy
Pierpaolo D’Urso, Rome, Italy
Gisella Facchinetti, Lecce, Italy
Peter Filzmoser, Vienna, Austria
Lluís Godo, Barcelona, Spain
Gil González Rodríguez, Oviedo, Spain
Mario Guarracino, Naples, Italy

vii
viii Organization

Janusz Kacprzyk, Warsaw, Poland

Frank Klawonn, Braunschweig/Wolfenbüttel, Germany
Rudolf Kruse, Magdeburg, Germany
Massimo Marinacci, Milan, Italy
Radko Mesiar, Bratislava, Slovakia
Enrique Miranda, Oviedo, Spain
Martin Stepnicka, Ostrava, Czech Republic
Wolfgang Trutschnig, Salzburg, Austria
Stefan Van Aelst, Ghent, Belgium

Publication Chair
Marek Gagolewski, Warsaw, Poland

Special Session Organizers

Patrizia Berti, Modena, Italy
Angela Blanco-Fernández, Oviedo, Italy
Andrea Capotorti, Perugia, Italy
Fabrizio Durante, Bolzano, Italy
Luis Angel García-Escudero, Valladolid, Spain
Brunero Liseo, Rome, Italy
Agustin Mayo-Iscar, Valladolid, Spain
Enrique Miranda, Oviedo, Spain
Niki Pfeifer, Munich, Germany
Ana B. Ramos-Guajardo, Oviedo, Spain
Giuseppe Sanfilippo, Palermo, Italy
Beatriz Sinova, Oviedo, Spain
Pedro Terán, Oviedo, Spain
Wolfgang Trutschnig, Salzburg, Austria

Additional Reviewers
Serena Arima, Salem Benferhat, Patrizia Berti, Angela Blanco-Fernández, Christian
Borgelt, Andrea Capotorti, Marco Cattaneo, Jasper de Bock, Alessandro De
Gregorio, Sara De la Rosa de Sáa, Marco Di Zio, Marta Disegna, Scott Ferson, Gianna
Figá-Talamanca, Enrico Foscolo, Marek Gagolewski, Luis Angel García-Escudero,
José Luis García-Lapresta, Angelo Gilio, Rita Giuliano, Stefania Gubbiotti,
Maria Letizia Guerra, Radim Jiroušek, Tomas Kroupa, Brunero Liseo, Massimo
Marinacci, Agustin Mayo-Iscar, Enrique Miranda, Piotr Nowak, David Over,
Sonia Pérez-Fernández, Davide Petturiti, Niki Pfeifer, Ana B. Ramos-Guajardo,
Organization ix

Pietro Rigo, Giuseppe Sanfilippo, Fabian Schmidt, Rudolf Seising, Beatriz Sinova,
Damjan Skulj, Yann Soullard, Andrea Tancredi, Luca Tardella, Pedro Terán,
Matthias Troffaes, Tiziana Tuoto, Arthur Van Camp, Paolo Vicig.

Local Organization
Marco Brandi
Massimiliano Coppola—conference website chair
Francesco Dotto
Liana Francescangeli—administrative staff
Tullia Padellini
Alessandra Pelorosso—administrative chair
Marco Stefanucci
Contents

Mean Value and Variance of Fuzzy Numbers with Non-continuous

Membership Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Luca Anzilli and Gisella Facchinetti
On the Construction of Radially Symmetric Trivariate Copulas . . . . . . 9
José De Jesús Arias García, Hans De Meyer and Bernard De Baets
Simulation of the Night Shift Solid Waste Collection System
of Phuket Municipality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Somsri Banditvilai and Mantira Niraso
Updating Context in the Equation: An Experimental
Argument with Eye Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Jean Baratgin, Brian Ocak, Hamid Bessaa
and Jean-Louis Stilgenbauer
Black-Litterman Model with Multiple Experts’ Linguistic Views . . . . . . 35
Marcin Bartkowiak and Aleksandra Rutkowska
Representing Lightweight Ontologies in a Product-Based
Possibility Theory Framework. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Salem Benferhat, Khaoula Boutouhami, Faiza Khellaf
and Farid Nouioua
Asymptotics of Predictive Distributions . . . . . . . . . . . . . . . . . . . . . . . . 53
Patrizia Berti, Luca Pratelli and Pietro Rigo
Independent k-Sample Equality Distribution Test
Based on the Fuzzy Representation . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Angela Blanco-Fernández and Ana B. Ramos-Guajardo
Agglomerative Fuzzy Clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Christian Borgelt and Rudolf Kruse

xi
xii Contents

Bayesian Inference for a Finite Population Total

Using Linked Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Dario Briscolini, Brunero Liseo and Andrea Tancredi
The Extension of Imprecise Probabilities Based on Generalized
Credal Sets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Andrey G. Bronevich and Igor N. Rozenberg
A Generalized SMART Fuzzy Disjunction of Volatility Indicators
Applied to Option Pricing in a Binomial Model . . . . . . . . . . . . . . . . . . 95
Andrea Capotorti and Gianna Figà-Talamanca
The Representation of Conglomerative Functionals . . . . . . . . . . . . . . . 103
Gianluca Cassese
The Likelihood Interpretation of Fuzzy Data . . . . . . . . . . . . . . . . . . . . 113
Marco E.G.V. Cattaneo
Combining the Information of Multiple Ranker in Ranked Set
Sampling with Fuzzy Set Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Bekir Cetintav, Selma Gurler, Neslihan Demirel and Gozde Ulutagay
A Savage-Like Representation Theorem for Preferences
on Multi-acts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Giulianella Coletti, Davide Petturiti and Barbara Vantaggi
On Some Functional Characterizations of (Fuzzy)
Set-Valued Random Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Ana Colubi and Gil Gonzalez-Rodriguez
Maximum Likelihood Under Incomplete Information:
Toward a Comparison of Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Inés Couso and Didier Dubois
The Use of Uncertainty to Choose Matching Variables
in Statistical Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Marcello D’Orazio, Marco Di Zio and Mauro Scanu
Beyond Fuzzy, Possibilistic and Rough: An Investigation
of Belief Functions in Clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Thierry Denœux and Orakanya Kanjanatarakul
Small Area Estimation in the Presence of Linkage Errors. . . . . . . . . . . 165
Loredana Di Consiglio and Tiziana Tuoto
A Test for Truncation Invariant Dependence . . . . . . . . . . . . . . . . . . . . 173
F. Marta L. Di Lascio, Fabrizio Durante and Piotr Jaworski
Contents xiii

Finite Mixture of Linear Regression Models: An Adaptive

Constrained Approach to Maximum Likelihood Estimation . . . . . . . . . 181
Roberto Di Mari, Roberto Rocci and Stefano Antonio Gattone
A Multivariate Analysis of Tourists’ Spending Behaviour . . . . . . . . . . . 187
Marta Disegna, Fabrizio Durante and Enrico Foscolo
Robust Fuzzy Clustering via Trimming and Constraints. . . . . . . . . . . . 197
Francesco Dotto, Alessio Farcomeni, Luis Angel García-Escudero
and Agustín Mayo-Iscar
One-Factor Lévy-Frailty Copulas with Inhomogeneous
Trigger Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Janina Engel, Matthias Scherer and Leonhard Spiegelberg
A Perceptron Classifier and Corresponding Probabilities . . . . . . . . . . . 213
Bernd-Jürgen Falkowski
Fuzzy Signals Fed to Gaussian Channels . . . . . . . . . . . . . . . . . . . . . . . 221
Laura Franzoi and Andrea Sgarro
Fuzzy Clustering Through Robust Factor Analyzers. . . . . . . . . . . . . . . 229
Luis Angel García-Escudero, Francesca Greselin
and Agustin Mayo Iscar
Consensus-Based Clustering in Numerical Decision-Making . . . . . . . . . 237
José Luis García-Lapresta and David Pérez-Román
Spatial Outlier Detection Using GAMs and Geographical
Information Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Alfonso García-Pérez and Yolanda Cabrero-Ortega
Centering and Compound Conditionals Under Coherence . . . . . . . . . . 253
Angelo Gilio, David E. Over, Niki Pfeifer and Giuseppe Sanfilippo
Approximate Bayesian Methods for Multivariate
and Conditional Copulae. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Clara Grazian and Brunero Liseo
The Sign Test for Interval-Valued Data . . . . . . . . . . . . . . . . . . . . . . . . 269
Przemysław Grzegorzewski and Martyna Śpiewak
Probability Distributions Related to Fuzzy P-Values . . . . . . . . . . . . . . . 277
Olgierd Hryniewicz
Probabilistic Semantics and Pragmatics for the Language
of Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Stefan Kaufmann
xiv Contents

Dynamic Analysis of the Development of Scientific Communities

in the Field of Soft Computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Ekaterina Kutynina and Alexander Lepskiy
Talk to Your Neighbour: A Belief Propagation Approach
to Data Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
Eleonora Laurenza
The Qualitative Characteristics of Combining Evidence
with Discounting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Alexander Lepskiy
Measuring the Dissimilarity Between the Distributions
of Two Random Fuzzy Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
María Asunción Lubiano, María Ángeles Gil, Beatriz Sinova,
María Rosa Casals and María Teresa López
An Empirical Analysis of the Coherence Between Fuzzy Rating
Scale- and Likert Scale-Based Responses to Questionnaires . . . . . . . . . 329
María Asunción Lubiano, Antonia Salas, Sara De la Rosa de Sáa,
Manuel Montenegro and María Ángeles Gil
Asymptotic Results for Sums of Independent Random Variables
with Alternating Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
Claudio Macci
Dispersion Measures and Multidistances on Rk . . . . . . . . . . . . . . . . . . 347
Javier Martín and Gaspar Mayor
Full Conglomerability, Continuity and Marginal Extension. . . . . . . . . . 355
Enrique Miranda and Marco Zaffalon
On Extreme Points of p-Boxes and Belief Functions . . . . . . . . . . . . . . . 363
Ignacio Montes and Sebastien Destercke
Modelling the Dependence in Multivariate Longitudinal Data
by Pair Copula Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
Marta Nai Ruscone and Silvia Angela Osmetti
Predictability in Probabilistic Discrete Event Systems . . . . . . . . . . . . . . 381
Farid Nouioua, Philippe Dague and Lina Ye
A Sandwich Theorem for Natural Extensions. . . . . . . . . . . . . . . . . . . . 391
Renato Pelessoni and Paolo Vicig
Envelopes of Joint Probabilities with Given Marginals
Under Absolute Continuity or Equivalence Constraints . . . . . . . . . . . . 399
Davide Petturiti and Barbara Vantaggi
Contents xv

Square of Opposition Under Coherence . . . . . . . . . . . . . . . . . . . . . . . . 407

Niki Pfeifer and Giuseppe Sanfilippo
Testing of Coarsening Mechanisms: Coarsening at Random
Versus Subgroup Independence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
Julia Plass, Marco E.G.V. Cattaneo, Georg Schollmeyer
and Thomas Augustin
Two-Sample Similarity Test for the Expected Value
of Random Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
Ana B. Ramos-Guajardo and Ángela Blanco-Fernández
Handling Uncertainty in Structural Equation Modeling . . . . . . . . . . . . 431
Rosaria Romano and Francesco Palumbo
Detecting Inconsistencies in Revision Problems. . . . . . . . . . . . . . . . . . . 439
Fabian Schmidt, Jörg Gebhardt and Rudolf Kruse
Tukey’s Biweight Loss Function for Fuzzy Set-Valued M-estimators
of Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
Beatriz Sinova and Stefan Van Aelst
Technical Gestures Recognition by Set-Valued Hidden Markov
Models with Prior Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
Yann Soullard, Alessandro Antonucci and Sébastien Destercke
Time Series Modeling Based on Fuzzy Transform . . . . . . . . . . . . . . . . 463
Luciano Stefanini, Laerte Sorini and Maria Letizia Guerra
Back to “Reasoning”. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
Marco Elio Tabacchi and Settimo Termini
Lexicographic Choice Functions Without Archimedeanicity . . . . . . . . . 479
Arthur Van Camp, Enrique Miranda and Gert de Cooman
Composition Operator for Credal Sets Reconsidered . . . . . . . . . . . . . . 487
Jiřina Vejnarová
A Nonparametric Linearity Test for a Multiple Regression Model
with Fuzzy Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
Dabuxilatu Wang
Treat a Weak Dependence Among Causes and Lives in Insurance
by Means of Fuzzy Sets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
Dabuxilatu Wang and Tengteng Wang
A Portfolio Diversification Strategy via Tail Dependence Clustering . . . 511
Hao Wang, Roberta Pappadà, Fabrizio Durante and Enrico Foscolo
xvi Contents

An Upper Bound Estimation About the Sample Average

of Interval-Valued Random Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
Xia Wang and Li Guan
On Asymptotic Properties of the Multiple Fuzzy Least Squares
Estimator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
Jin Hee Yoon, Seung Hoe Choi and Przemyslaw Grzegorzewski

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533

Mean Value and Variance of Fuzzy Numbers
with Non-continuous Membership Functions

Luca Anzilli and Gisella Facchinetti

Abstract We propose a definition of mean value and variance for fuzzy numbers
whose membership functions are upper-semicontinuous but are not necessarily con-
tinuous. Our proposal uses the total variation of bounded variation functions.

1 Introduction

In this work we face the problem of defining the mean value and variance of fuzzy
numbers whose membership functions are upper-semicontinuous but are not neces-
sarily continuous. The literature presents a high number of definitions in the con-
tinuous case [3–6, 9] also because the average has often been counted among the
ranking modes of fuzzy numbers. The starting point is that a fuzzy number is defined
by a membership function that is of bounded variation. Even in this more general
context, it is possible to introduce a weighted average by means of a classical vari-
ational formulation and by α-cuts, as well as many authors do in the continuous
case. In the non-continuous case, we introduce the lower and upper weighted mean
values, but the generality of the weights doesn’t let to obtain the weighted mean
value as the middle point of the previous ones. This property will be obtained either
in the continuous case or for particular weight functions. An interesting property of
this new version is connected to the view of the weighted mean value in a possi-
bilistic framework, as in the continuous case Carlsson and Fullér [3] and Fullér and
Majlender [5] do. This property is interesting and harbinger of future developments
also in the non-continuous case. Following the same line, we pass to introduce the
concept of variance and we suggest two different definitions as happens in the case
of continuous membership functions.

L. Anzilli (B) · G. Facchinetti

Department of Management, Economics, Mathematics and Statistics,
University of Salento, Lecce, Italy
e-mail: luca.anzilli@unisalento.it
G. Facchinetti
e-mail: gisella.facchinetti@unisalento.it

© Springer International Publishing Switzerland 2017 1

M.B. Ferraro et al. (eds.), Soft Methods for Data Science, Advances
in Intelligent Systems and Computing 456, DOI 10.1007/978-3-319-42972-4_1
2 L. Anzilli and G. Facchinetti

2 Bounded Variation Functions

In this section we recall some basic properties of functions of bounded variation. For
more details see, e.g., [7].
Let I ⊂ R be an interval and u : I → R be a function. The total variation of u on

n
I is defined as V I [u] = sup |u(xi ) − u(xi−1 )|, where the supremum is taken over
i=1
all finite partitions {x0 , x1 , . . . , xn } ⊂ I with x0 < x1 < · · · < xn . We say that u is a
bounded variation function on I if V I [u] < +∞. We denote BV (I ) the space of all
bounded variation functions on I . The following properties hold: if u is monotone
and bounded on I then u ∈ BV (I ) and V I [u] = sup I u − inf I u; if u 1 , u 2 ∈ BV (I )
and k is a constant then ku 1 , u 1 + u 2 , u 1 − u 2 , u 1 u 2 belong to BV (I ).
Let u : R → R be a bounded variation function. The total variation function of u is
the increasing function vu defined by vu (x) = V−∞ x
[u], where V−∞
x
[u] denotes the
total variation of u on ] − ∞, x].
The total variation measure |Du| of u is defined as the Lebesgue-Stieltjes measure
dvu associated to vu . The positive and negative variations of u are defined, respec-
tively, by the increasing functions

u + (x) = (vu (x) + u(x)) /2 , u − (x) = (vu (x) − u(x)) /2 . (1)

We have u = u + − u − , the so-called Jordan decomposition of u. The integral of a

measurable function g with respect to |Du| is given by
  
g(x) |Du| = g(x) du + (x) + g(x) du − (x) (2)

where du + and du − are the Lebesgue-Stieltjes measures associated to u + and u − ,

respectively. In general, the total variation measure |Du| is not absolutely continuous
with respect
 to Lebesgue
 measure. However, if u is an absolutely continuous function
then g(x) |Du| = g(x) |u  (x)|d x.

3 Weighted Mean Value of Fuzzy Numbers

In this section we propose a definition of f -weighted mean value for a fuzzy number
whose membership function is upper-semicontinuous but not necessarily continuous.
These properties produce that the fuzzy number has membership function of bounded
variation and so we realize to introduce its weighted mean value by its total variation
measure.
A fuzzy number A is a fuzzy set of R with a normal, (fuzzy) convex and upper-
semicontinuous membership function μ : R → [0, 1] of bounded support [2]. From
Mean Value and Variance of Fuzzy Numbers … 3

the previous definition there exist two functions μ L , μ R : R → [0, 1], where μ L is
nondecreasing and right-continuous and μ R is nonincreasing and left-continuous,
such that ⎧
⎪
⎪ 0 x < a1 or x > a4
⎨
μ L (x) a1 ≤ x < a2
μ(x) =
⎪
⎪ 1 a2 ≤ x ≤ a3
⎩
μ R (x) a3 < x ≤ a4

with a1 ≤ a2 ≤ a3 ≤ a4 . The α-cut of a fuzzy set A, 0 ≤ α ≤ 1, is the crisp set

Aα = {x ∈ X ; μ(x) ≥ α} if 0 < α ≤ 1 and A0 = [a1 , a4 ] if α = 0. Every α-cut of
a fuzzy number is a closed interval Aα = [a L (α), a R (α)].
We observe that the membership function μ of a fuzzy number A is a bounded
variation function. We propose a definition of mean value of A using the total variation
1
measure |Dμ| of μ. Let f be a weighting function such that f > 0 and 0 f (α) dα =
1.
Definition 1 We define the f -weighted mean value of a fuzzy number A as
 +∞  +∞
E(A; f ) = x f (μ(x)) |Dμ| f (μ(x)) |Dμ| . (3)
−∞ −∞

In order to define the lower and upper mean values of A, we consider the positive
and negative variations of μ, as defined in (1), given by, respectively,
⎧ ⎧
⎪
⎨0 x < a1 ⎪
⎨0 x ≤ a3
+ −
μ (x) = μ L (x) a1 ≤ x < a2 μ (x) = 1 − μ R (x) a3 < x ≤ a4 (4)
⎪
⎩ ⎪
⎩
1 x ≥ a2 , 1 x > a4 .

We observe that μ+ is an increasing and right continuous function and that μ− is an

increasing and left continuous function.
Definition 2 We define the lower and upper f -weighted mean values of A, respec-
tively, as
 +∞ +
 +∞
−∞ x f (μ(x)) dμ (x) x f (μ(x)) dμ− (x)
E ∗ (A; f ) =  +∞ , E (A; f ) = −∞
∗
+∞ .
+ −
−∞ f (μ(x)) dμ (x) −∞ f (μ(x)) dμ (x)
(5)
From previous definition and (2) we deduce the following result
Proposition 1 We have E(A; f ) = (1 − w)E ∗(A; f ) + wE ∗ (A; f ), where w ∈
+∞ +∞
(0, 1) is defined as w = −∞ f (μ(x)) dμ− (x) −∞ f (μ(x)) |Dμ|.
4 L. Anzilli and G. Facchinetti

3.1 Weighted Mean Value Using α-cuts

Following the line present in the classical case of continuous fuzzy numbers, we face
the problem to give an expression of lower and upper f -weighted mean values using
α-cuts. This different version lets the possibility to write the previous definitions by
the left and right extreme points of fuzzy number α-cuts, exactly as happens in the
continuous case.
Proposition 2
1 1
a L (α) f (μ(a L (α))) dα a R (α) f (μ(a R (α))) dα
E ∗ (A; f ) = 0  1 , E ∗ (A; f ) = 0  1 . (6)
0 f (μ(a L (α))) dα 0 f (μ(a R (α))) dα

Proof First, we prove the following equalities:

 +∞  1
+
g(x) f (μ(x)) dμ (x) = g(a L (α)) f (μ(a L (α))) dα , (7)
−∞ 0
 +∞  1
g(x) f (μ(x)) dμ− (x) = g(a R (α)) f (μ(a R (α))) dα . (8)
−∞ 0

Since a L (α) = min{x ≥ a1 ; μ L (x) ≥ α}, α ∈ [0, 1], applying the change of vari-
+∞ a2
able formula [7, Theorem 5.42] we obtain g(x) f (μ(x)) dμ+ (x) = g(x)
−∞
1 a1
f (μ(x)) dμ L (x) = 0 g(a L (α)) f (μ(a L (α))) dα and thus (7). We now prove (8).
 +∞  a3  −a3
We have g(x) f (μ(x)) dμ− (x) = g(x) f (μ(x)) dμ R (x) = g(−y)
−∞  −a3 a4 −a4
f (μ(−y)) dμ R (−y) = −a4 g(−y) f (μ(−y)) dk(y)
where we have used the change of variable y = −x and let k(x) = μ R (−x).
Since a R (α) = max{x; x ≤ a4 , μ R (x) ≥ α}, α ∈ [0, 1], letting h(α) = −a R (α) =
min{y; y ≥ −a4 , k(y) ≥ α} and using the change of variable formula [7, Theorem
5.42] (observing that k is increasing), we have, continuing the above chain of equal-
ities,
 
k(−a3 ) μ R (a3 )
= g(−h(α)) f (μ(−h(α))) dα = g(a R (α)) f (μ(a R (α))) dα =
k(−a4 ) μ R (a4 )
1
0 g(a R (α)) f (μ(a R (α))) dα. Then (8) is proved. Substituting (7) and (8) in (5)
with g(x) = x and g(x) = 1 we obtain for E ∗ (A; f ) and E ∗ (A; f ) the expressions
given in (6).

We observe that w of Proposition 1 can be expressed in terms of α-cuts as

1 1 1
w = 0 f (μ(a R (α))) dα 0 f (μ(a L (α))) dα + 0 f (μ(a R (α))) dα .
Mean Value and Variance of Fuzzy Numbers … 5

3.2 The Case f (α) = 1

We now consider the special case f (α) = 1. We denote by E(A), E ∗ (A) and E ∗ (A),
respectively, the mean value and lower and upper mean values of a fuzzy number A
computed using f (α) = 1. This particular case is interesting as we may show that
a sufficient condition so that the weight w is equal to 1/2 is that f (α) = 1. This
condition is not necessary as we will show later in an example.
 +∞ 1
Proposition 3 We have E(A) = 21 −∞ x |Dμ| = 21 0 (a L (α) + a R (α)) dα,
 +∞ 1  +∞ 1
E ∗ (A) = −∞ x dμ+ (x) = 0 a L (α) dα and E ∗ (A) = −∞ x dμ− (x) = 0 a R (α)
dα. In particular E(A) = (E ∗ (A) + E ∗ (A))/2.
 +∞ +∞ +
Proof Since μ+ (x) is increasing, we have (see Sect. 2) −∞ dμ+ (x) = V−∞ [μ ] =
+ +
 +∞
sup μ − inf μ = 1, where last equality follows from (4). Similarly we get −∞
 +∞  +∞  +∞
dμ− (x) = 1. Then, from (2), −∞ |Dμ| = −∞ dμ+ (x) + −∞ dμ− (x) = 2. Sub-
stituting in (3), with f = 1, we have the first equality of the thesis for E(A). The
second equality follows observing that w, as defined in Proposition 1, is equal to 1/2
and using (6) in Proposition 1 with f (α) = 1.

3.3 Example

We now present an example to show that, when f (α) = 1, we may have w = 1/2.
We compute the mean values of two fuzzy numbers shown in Fig. 1 for f (α) = 2α.
 +∞
First, we consider the fuzzy number (a). We have −∞ x f (μ(x)) dμ+ (x) = 25/12,
 +∞ +
 +∞
−∞ f (μ(x)) dμ (x) = 5/4 and thus E ∗ (A; f ) = 5/3. Furthermore, −∞ x
+∞
f (μ(x)) dμ− (x) = 59/16, −∞ f (μ(x)) dμ− (x) = 17/16 and then E ∗ (A; f ) =
59/17. Moreover E(A; f ) = 277/111 ≈ 2.5 and w = 17/37  +∞ = 1/2.
We now consider the fuzzy number (b). We have −∞ x f (μ(x)) dμ+ (x) =
 +∞  +∞
25/12, −∞ f (μ(x)) dμ+ (x) = 5/4 and thus E ∗ (A; f ) = 5/3. Furthermore, −∞
 +∞
x f (μ(x)) dμ− (x) = 55/12, −∞ f (μ(x)) dμ− (x) = 5/4 and then E ∗ (A; f ) =
11/3. Moreover E(A; f ) = 8/3 ≈ 2.7 and w = 1/2.

3.4 Possibilistic Framework

We observe that our proposal, which starts from a variational measure, may be viewed
in a possibilistic framework. Indeed, since Poss(A ≤ a L (α)) = supx≤aL (α) μ(x) =
μ(a L (α)) and Poss(A ≥ a R (α)) = supx≥a R (α) μ(x) = μ(a R (α)), from (6) we get
1 1
E ∗ (A; f ) = 0 a L (α) f (Poss(A ≤ a L (α))) dα 0 f (Poss(A ≤ a L (α))) dα ,
6 L. Anzilli and G. Facchinetti

(a) (b)
1 1

1/2 1/2

1/4

1 1 3 4
3 4

Fig. 1 Two fuzzy numbers with non-continuous membership function

1 1
E ∗ (A; f ) = 0 a R (α) f (Poss(A ≥ a R (α))) dα 0 f (Poss(A ≥ a R (α))) dα.
Note that A does not need to be continuous. In the special case when A is a continuous
fuzzy number we retrieve the lower and upper possibilistic mean values proposed
in [3] and [5] (see Proposition 4). Thus our approach offers an extension of possi-
bilistic mean values to the case of fuzzy numbers whose membership functions are
upper-semicontinuous but are not necessarily continuous.

3.5 Weighted Mean Value of Continuous Fuzzy Numbers

If we use the variational approach for continuous fuzzy numbers we obtain that the
weighted mean value is the simple average of upper and lower f -weighted mean
values. This means that for every weighting function f the weight w = 1/2.

 1 continuous membership function μ then

Proposition 4 If A is a fuzzy number with
+∞
E(A; f ) = 21 −∞ x f (μ(x)) |Dμ| = 21 0 (a L (α) + a R (α)) f (α) dα, and E ∗ (A
 +∞ 1  +∞
; f ) = −∞ x f (μ(x)) dμ+ (x) = 0 a L (α) f (α) dα, E ∗ (A; f ) = −∞ x f (μ(x))

dμ− (x) = 0 a R (α) f (α) dα. Moreover, w = 1/2 and thus E(A; f ) = (E ∗ (A; f ) +
1
∗
E (A; f ))/2.

Proof The assertions easily follow from previous results observing that, since μ
is continuous, we have μ(a L (α)) = α and μ(a R (α)) = α. In particular, we have
 +∞ 1 1
f (μ(x)) dμ+ (x) = 0 f (μ(a L (α))) dα = 0 f (α) dα = 1 and, similarly,
−∞
+∞ −
−∞ f (μ(x)) dμ (x) = 1.
Mean Value and Variance of Fuzzy Numbers … 7

4 Weighted Variance of Fuzzy Numbers

In this section we introduce two definitions of f -weighted variance of a fuzzy num-

ber A whose membership function is upper semicontinuous. The first, V ar1 (A), is
obtained as the simple average of the lower and upper variances of A, that derive
from the definition of lower and upper mean value of A given in Definition 2. The
second, V ar2 (A), is the natural definition of variance that starts from Definition 1.
In the continuous case and for particular weights f we recover classical definitions
introduced by other authors.
Definition 3 We define the lower and upper variances of a, respectively, as
 +∞  +∞
V ar∗ (A; f ) = (x − E ∗ (A; f ))2 f (μ(x)) dμ+ (x) f (μ(x)) dμ+ (x) ,
−∞ −∞
 +∞  +∞
2
V ar ∗ (A; f ) = x − E ∗ (A; f ) f (μ(x)) dμ− (x) f (μ(x)) dμ− (x) .
−∞ −∞

and the variance of A as V ar1 (A; f ) = (V ar∗ (A; f ) + V ar ∗ (A; f ))/2.

Using (7) and (8) with g(x) = (x − E ∗ (A; f ))2 and g(x) = (x − E ∗ (A; f ))2 ,
respectively, we may express previous definitions in terms of α-cuts as follows
 1  1
V ar∗ (A; f ) = (a L (α) − E ∗ (A; f ))2 f (μ(a L (α))) dα f (μ(a L (α))) dα,
0 0
 
1 2 1
V ar ∗ (A; f ) = a R (α) − E ∗ (A; f ) f (μ(a R (α))) dα f (μ(a L (α))) dα.
0 0

Definition 4 Alternatively, we may define the variance of a fuzzy number A as

 +∞  +∞
V ar2 (A; f ) = −∞ (x − E(A; f ))2 f (μ(x)) |Dμ| −∞ f (μ(x))|Dμ|.

4.1 Weighted Variance of Continuous Fuzzy Numbers

Proposition 5 If A is a fuzzy number with continuous membership function μ then

1
V ar ∗ (A; f ) = 0
1
V ar∗ (A; f ) = 0 (a L (α) − E ∗ (A; f ))2 f (α) dα and
(a R (α) − E ∗ (A; f ))2 f (α) dα.

Remark 1 If A is a continuous fuzzy number and f (α) = 2α we obtain that V ar1 (A)
matches the variance proposed by Zhang and Nie [9].

Proposition 6 If A is a continuous fuzzy number then 

1
V ar2 (A; f ) = 21 0 (a L (α) − E(A; f ))2 + (a R (α) − E(A; f ))2 f (α) dα.
8 L. Anzilli and G. Facchinetti

Remark 2 If A is a continuous fuzzy number and f (α) = 2α we obtain that

V ar2 (A; f ) agrees with the variance V ar  (A) introduced by Carlsson and Fullér
[3].

Proposition 7 If A is a fuzzy number with differentiable membership function μ then

+∞
V ar2 (A; f ) = 21 −∞ (x − E(A; f ))2 f (μ(x)) |μ (x)| d x.

Remark 3 If A is a fuzzy number with differentiable membership function μ and

f (α) = 2α then V ar2 (A) is the variance proposed by Li, Guo and Yu [8].

5 Conclusion

In this paper we have presented the weighted mean value and variance definitions for
fuzzy numbers with upper semicontinuous membership functions that are of bounded
variation. In the same context in a previous paper we have looked for a definition of
evaluation and ranking for fuzzy numbers [1]. The idea to use the space of bounded
variation functions has allowed us to view the mean value and the variance of fuzzy
numbers either in a variational context or in a classical way by α-cuts. This choice,
thanks to the freedom that we have left to the weights, has shown interesting results
that are able to generalize the previous ones but even to recover others present in
literature for the continuous case. We have intention to continue in this direction to
see if the space of bounded variation functions offers a new research field for fuzzy
sets.

References

1. Anzilli L, Facchinetti G (2013) The total variation of bounded variation functions to evaluate
and rank fuzzy quantities. Int J Intell Syst 28(10):927–956
2. Bede B (2013) Mathematics of fuzzy sets and fuzzy logic. Springer
3. Carlsson C, Fullér R (2001) On possibilistic mean value and variance of fuzzy numbers. Fuzzy
Sets Syst 122(2):315–326
4. Dubois D, Prade H (1987) The mean value of a fuzzy number. Fuzzy Sets Syst 24(3):279–300
5. Fullér R, Majlender P (2003) On weighted possibilistic mean and variance of fuzzy numbers.
Fuzzy Sets Syst 136(3):363–374
6. Georgescu I (2009) Possibilistic risk aversion. Fuzzy Sets Syst 160(18):2608–2619
7. Leoni G (2009) A first course in Sobolev spaces, vol 105. American Mathematical Society
Providence, RI
8. Li X, Guo S, Yu L (2015) Skewness of fuzzy numbers and its applications in portfolio selection.
Fuzzy Syst IEEE Trans 23(6):2135–2143
9. Zhang WG, Nie ZK (2003) On possibilistic variance of fuzzy numbers. In: Rough sets, fuzzy
sets, data mining, and granular computing.Springer, pp 398–402
On the Construction of Radially Symmetric
Trivariate Copulas

José De Jesús Arias García, Hans De Meyer and Bernard De Baets

Abstract We propose a method to construct a 3-dimensional symmetric function

that is radially symmetric, using two symmetric 2-copulas, with one of them being
also radially symmetric. We study the properties of the presented construction in
some specific cases and provide several examples for different families of copulas.

Keywords Copula · Quasi-copula · Radial symmetry · Aggregation function

1 Introduction

An n-dimensional copula (or, for short, n-copula) is a multivariate distribution func-

tion with the property that all its n univariate marginals are uniform distributions on
[0, 1]. Formally, an n-copula is a [0, 1]n → [0, 1] function that satisfies the following
conditions:
1. Cn (x) = 0 if x is such that x j = 0 for some j ∈ {1, 2, . . . , n}.
2. Cn (x) = x j if x is such that xi = 1 for all i = j.
3. Cn is n-increasing, i.e., for any n-box P = [a1 , b1 ] × · · · × [an , bn ] ⊆ [0, 1]n it
holds that 
VCn (P) = (−1) S(x) Cn (x) ≥ 0 ,
x∈vertices(P)

J.D.J. Arias García (B) · B. De Baets

KERMIT, Department of Mathematical Modelling, Statistics and Bioinformatics, Ghent
University, Coupure links 653, 9000 Ghent, Belgium
e-mail: JoseDeJesus.AriasGarcia@ugent.be
B. De Baets
e-mail: Bernard.DeBaets@ugent.be
H. De Meyer
Department of Applied Mathematics and Computer Science, Ghent University,
Krijgslaan 281 S9, 9000 Ghent, Belgium
e-mail: Hans.DeMeyer@ugent.be

© Springer International Publishing Switzerland 2017 9

M.B. Ferraro et al. (eds.), Soft Methods for Data Science, Advances
in Intelligent Systems and Computing 456, DOI 10.1007/978-3-319-42972-4_2
10 J.D.J. Arias García et al.

where S(x) = #{ j ∈ {1, 2, . . . , n} | x j = a j }. VCn (P) is called the Cn -volume

of P.
Due to Sklar’s theorem, which states that any continuous multivariate distribution
function can be written in terms of its n univariate marginals by means of a unique
n-copula, n-copulas have become one of the most important tools for the study of
certain types of properties of random vectors, such as stochastic dependence (see [4,
13] for more details on n-copulas). One example of a property that can be directly
studied from copulas is the property of radial symmetry. An n-dimensional random
vector (X 1 , . . . , X n ) is said to be radially symmetric about (x1 , . . . , xn ) if the distri-
bution of the random vector (X 1 − x1 , . . . , X n − xn ) is the same as the distribution
of the random vector (x1 − X 1 , . . . , xn − X n ). It is easily shown that radial sym-
metry can be characterized using copulas: a random vector (X 1 , . . . , X n ) is radially
symmetric about (x1 , . . . , xn ) if and only if for any j ∈ {1, . . . , n} X j − x j has the
same distribution as x j − X j and the n-copula Cn associated to the random vector
satisfies the identity Cn = Ĉn , where Ĉn denotes the survival n-copula associated to
Cn , and which can be computed as:


n 
n
Ĉn (x1 , . . . , xn ) = x j − (n − 1) + Cn (1, . . . , 1 − xi , . . . , 1 − x j , . . . , 1)
j=1 i< j

n
− Cn (1, . . . , 1 − xi , . . . , 1 − x j , . . . 1 − xk , . . . , 1) + . . .
i< j<k
+(−1) Cn (1 − x1 , 1 − x2 , . . . , 1 − xn ) .
n
(1)

Due to this characterization, we say that an n-copula Cn is radially symmetric if it

satisfies the identity Cn = Ĉn . Survival copulas also have a probabilistic interpreta-
tion. If the random vector (X 1 , . . . X n ) has the copula Cn as its distribution function,
then Ĉn is the distribution function of the random vector (1 − X 1 , . . . , 1 − X n ). This
probabilistic interpretation has led to several studies of the transformations of copu-
las which are induced by certain types of transformations on random variables (see
[6–8]). These transformations have been generalized and studied in the framework
of aggregation functions [1–3].
Radially symmetric copulas have a particular importance in stochastic simulation,
as they are used as a part of the multivariate version of the antithetic variates method,
which is a variance reduction technique used in Monte Carlo methods (see [12]).
In the binary case, well-known examples of families of bivariate copulas that are
radially symmetric are the Gaussian family, the Frank family and the Farlie-Gumbel-
Morgenstern (FGM) family. However, there are few attempts to construct families of
n-copulas with some specific properties for n ≥ 3. In this contribution we propose a
construct method for trivariate radially symmetric copulas.
On the Construction of Radially Symmetric Trivariate Copulas 11

2 Radial Symmetry and Associativity

Archimedean n(-quasi)-copulas are one of the most well-known classes of n-copulas,

which have the additional property that they are symmetric, i.e., for any permutation
σ of {1, 2, . . . , n} and for any x ∈ [0, 1]n it holds that

Cn (x1 , x2 , . . . , xn ) = Cn (xσ (1) , xσ (2) , . . . , xσ (n) ) .

Archimedean n(-quasi)-copulas are also associative and can be defined recur-

sively, i.e., for any x, y, z ∈ [0, 1] the equality C2 (x, C2 (y, z)) = C2 (C2 (x, y), z)
holds and for any n ≥ 2 and x ∈ [0, 1]n+1 it holds that

Cn+1 (x) = C2 (x1 , Cn (x2 , . . . , xn+1 ))

= C2 (Cn (x1 , . . . , xn ), xn+1 ) .

It is well known that Archimedean n(-quasi)-copulas can be fully character-

ized in terms of an additive generator (see [11]). Some further generalizations of
Archimedean copulas have been proposed; for example in [10], nested Archimedean
n-copulas are studied, where several bivariate Archimedean copulas are iterated to
construct an n-copula (for example, in the trivariate case C2 (x, D2 (y, z)) would be
an example of such a construction, where C2 and D2 are bivariate Archimedean
copulas).
If a 2-copula is radially symmetric, then for any (x, y) ∈ [0, 1]2 , it holds that

C2 (x, y) + (1 − C2 (1 − x, 1 − y)) = x + y .

If C2 is an Archimedean copula, the latter equation is a particular case of a func-

tional equation studied by Frank in [5]. More specifically in [5], it is proven that if a
continuous function F : [0, 1]2 → [0, 1] satisfies the following properties
1. for any x ∈ [0, 1] it holds that F(x, 0) = F(0, x) = 0,
2. for any x ∈ [0, 1] it holds that F(x, 1) = F(1, x) = x,
3. the functions F(x, y) and G(x, y) = x + y − F(x, y) are both associative;
then F must be a member of the Frank family of copulas or an ordinal sum constructed
from members of this family. The Frank family is given by:
 
(α) 1 (e−αx − 1)(e−αy − 1)
F (x, y) = − ln 1 + ,
α e−α − 1

where α ∈ R ∪ {−∞, ∞} (Although the case α = ∞ is not an Archimedean copula).

In [8] this result is complemented by showing that 2-copulas that are both asso-
ciative and radially symmetric are members of the Frank family of copulas or an
(α )
ordinal sum of the form C2 = ( a j , b j , F( j)j ) j∈J , such that for any j, there exists
i j with the property that α j = αi j , a j = 1 − bi j and b j = 1 − ai j .
12 J.D.J. Arias García et al.

Note that if a 3-copula is radially symmetric, then its 2-dimensional marginals

must also be radially symmetric. This trivially generalizes to higher dimensions.
Hence if an associative 3-copula is radially symmetric, then its 2-dimensional mar-
ginals must also be solutions of the Frank functional equation. Unfortunately, as
shown in [9], for n ≥ 3 the only solutions are the product copula Πn (x1 , . . . , xn ) =
x1 x2 . . . xn (which is the copula of independent random variables) and the minimum
operator Mn (x1 , . . . , xn ) = min(x1 , . . . , xn ) (which is the copula of comonotonic
random variables) or ordinal sums constructed using these two n-copulas. From this
it follows that if we want to construct radially symmetric copulas in higher dimen-
sions, we must weaken the condition of associativity (and as a consequence the
Archimedean property is lost), as jointly requiring both properties is too restrictive
in higher dimensions.

3 The Construction

In [8], the authors study several transformations of bivariate copulas. They show that
every radially symmetric 2-copula has the following form

C2 (x, y) + Ĉ2 (x, y)

, (2)
2
where C2 is a 2-copula. This result is easily generalized to higher dimensions. Keep-
ing this result in mind, we propose the following construction method in three dimen-
sions.
Definition 1 Let C2 , D2 be two symmetric 2-copulas, such that C2 is also radially
symmetric. We define the symmetric function SC2 ,D2 : [0, 1]3 → R associated to the
pair C2 , D2 as

1
SC2 ,D2 (x, y, z) = [1 − x − y − z + C2 (x, y) + C2 (x, z) + C2 (y, z)]
2
1
+ [HC2 ,D2 (x, y, z) − HC2 ,D2 (1 − x, 1 − y, 1 − z)] , (3)
2
where

HC2 ,D2 (x, y, z) = D2 (x, C2 (y, z)) + D2 (y, C2 (x, z)) + D2 (z, C2 (x, y))
2
− [D2 (x, D2 (y, z)) + D2 (y, D2 (x, z)) + D2 (z, D2 (x, y))] .
3
Note that the function HC2 ,D2 satisfies the boundary conditions of a 3-copula,
and from this is easy to prove that SC2 ,D2 also satisfies the boundary conditions of a
copula, and that the bivariate marginals of SC2 ,D2 are all equal to C2 .
On the Construction of Radially Symmetric Trivariate Copulas 13

Proposition 1 Let C2 , D2 be two symmetric 2-copulas, such that C2 is also radially

symmetric. Let SC2 ,D2 be defined as in Eq. (3). If SC2 ,D2 is a 3-copula, then it is radially
symmetric.

Proof It can be shown after some tedious computations that we can rewrite SC2 ,D2
as

SC2 ,D2 (x, y, z) = x + y + z − 2 + SC2 ,D2 (1 − x, 1 − y, 1)

+SC2 ,D2 (1 − x, 1, 1 − z) + SC2 ,D2 (1, 1 − y, 1 − z)
−SC2 ,D2 (1 − x, 1 − y, 1 − z) ,

i.e., from the definition of survival copula given by Eq. (1), it follows that if SC2 ,D2
is a 3-copula, then SC2 ,D2 is a radially symmetric 3-copula because it coincides with
its associated survival copula. 

However, SC2 ,D2 is not necessarily a 3-copula, as it may even not be an increas-
ing function. For example, if C2 = D2 = F (−2) , we can easily see that S F (−2) ,F (−2)
( 21 , 10
1
, 10
1
) < 0 = S F (−2) ,F (−2) (0, 10
1
, 10
1
). We now provide some examples where the
construction effectively yields a 3-copula.
Example 1 Consider the Frank family of 2-copulas. From the results in [11], we
know that for α ≥ − ln(2), the 3-dimensional version of the Frank 2-copula, given by
F3(α) (x, y, z) = F (α) (x, F (α) (y, z)), is a 3-copula. From this, it follows immediately
that if α ≥ − ln(2) then S F (α) ,F (α) is a 3-copula. However, with some computational
help, it can be shown that S F (α) ,F (α) is also a 3-copula for α ≥ − ln(3).

Example 2 The FGM family of bivariate copulas is given by

F (θ) (x, y) = x y + θ x y(1 − x)(1 − y) , θ ∈ [−1, 1] .

In this case, some

√ computations
√ show that S F (θ ) ,F (θ ) is a 3-copula if and only if
θ ∈ [−1/2(3 − 5), 1/2( 21 − 3)].

Example 3 For any 2-copula C2 , SC2 ,Π2 is a 3-copula if and only if for any
x1 , x2 , y1 , y2 , z 1 , z 2 ∈ [0, 1] such that x1 ≤ x2 , y1 ≤ y2 , z 1 ≤ z 2 it holds that

(x2 − x1 )VC2 ([y1 , y2 ] × [z 1 , z 2 ]) + (y2 − y1 )VC2 ([x1 , x2 ] × [z 1 , z 2 ])

+(z 2 − z 1 )VC2 ([x1 , x2 ] × [y1 , y2 ])
≥ 2(x2 − x1 )(y2 − y1 )(z 2 − z 1 ) .

If C2 is absolutely continuous, then this last condition is equivalent to:

∂ 2 C2 ∂ 2 C2 ∂ 2 C2
(x, y) + (x, z) + (y, z) ≥ 2 . (4)
∂ x∂ y ∂ x∂z ∂ y∂z
Random documents with unrelated
content Scribd suggests to you:
Gutenberg” appears, or with which the phrase “Project
Gutenberg” is associated) is accessed, displayed, performed,
viewed, copied or distributed:

This eBook is for the use of anyone anywhere in the United

States and most other parts of the world at no cost and
with almost no restrictions whatsoever. You may copy it,
give it away or re-use it under the terms of the Project
Gutenberg License included with this eBook or online at
www.gutenberg.org. If you are not located in the United
States, you will have to check the laws of the country
where you are located before using this eBook.

1.E.2. If an individual Project Gutenberg™ electronic work is

derived from texts not protected by U.S. copyright law (does not
contain a notice indicating that it is posted with permission of
the copyright holder), the work can be copied and distributed to
anyone in the United States without paying any fees or charges.
If you are redistributing or providing access to a work with the
phrase “Project Gutenberg” associated with or appearing on the
work, you must comply either with the requirements of
paragraphs 1.E.1 through 1.E.7 or obtain permission for the use
of the work and the Project Gutenberg™ trademark as set forth
in paragraphs 1.E.8 or 1.E.9.

1.E.3. If an individual Project Gutenberg™ electronic work is

posted with the permission of the copyright holder, your use and
distribution must comply with both paragraphs 1.E.1 through
1.E.7 and any additional terms imposed by the copyright holder.
Additional terms will be linked to the Project Gutenberg™
License for all works posted with the permission of the copyright
holder found at the beginning of this work.

1.E.4. Do not unlink or detach or remove the full Project

Gutenberg™ License terms from this work, or any files
 containing a part of this work or any other work associated with
Project Gutenberg™.

1.E.5. Do not copy, display, perform, distribute or redistribute

this electronic work, or any part of this electronic work, without
prominently displaying the sentence set forth in paragraph 1.E.1
with active links or immediate access to the full terms of the
Project Gutenberg™ License.

1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if
you provide access to or distribute copies of a Project
Gutenberg™ work in a format other than “Plain Vanilla ASCII” or
other format used in the official version posted on the official
Project Gutenberg™ website (www.gutenberg.org), you must,
at no additional cost, fee or expense to the user, provide a copy,
a means of exporting a copy, or a means of obtaining a copy
upon request, of the work in its original “Plain Vanilla ASCII” or
other form. Any alternate format must include the full Project
Gutenberg™ License as specified in paragraph 1.E.1.

1.E.7. Do not charge a fee for access to, viewing, displaying,

performing, copying or distributing any Project Gutenberg™
works unless you comply with paragraph 1.E.8 or 1.E.9.

1.E.8. You may charge a reasonable fee for copies of or

providing access to or distributing Project Gutenberg™
electronic works provided that:

• You pay a royalty fee of 20% of the gross profits you derive
from the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
 payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”

• You provide a full refund of any money paid by a user who

notifies you in writing (or by e-mail) within 30 days of receipt
that s/he does not agree to the terms of the full Project
Gutenberg™ License. You must require such a user to return or
destroy all copies of the works possessed in a physical medium
and discontinue all use of and all access to other copies of
Project Gutenberg™ works.

• You provide, in accordance with paragraph 1.F.3, a full refund of

any money paid for a work or a replacement copy, if a defect in
the electronic work is discovered and reported to you within 90
days of receipt of the work.

• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.

1.E.9. If you wish to charge a fee or distribute a Project

Gutenberg™ electronic work or group of works on different
terms than are set forth in this agreement, you must obtain
permission in writing from the Project Gutenberg Literary
Archive Foundation, the manager of the Project Gutenberg™
trademark. Contact the Foundation as set forth in Section 3
below.

1.F.

1.F.1. Project Gutenberg volunteers and employees expend

considerable effort to identify, do copyright research on,
transcribe and proofread works not protected by U.S. copyright
law in creating the Project Gutenberg™ collection. Despite these
efforts, Project Gutenberg™ electronic works, and the medium
on which they may be stored, may contain “Defects,” such as,
but not limited to, incomplete, inaccurate or corrupt data,
transcription errors, a copyright or other intellectual property
infringement, a defective or damaged disk or other medium, a
computer virus, or computer codes that damage or cannot be
read by your equipment.

1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES - Except

for the “Right of Replacement or Refund” described in
paragraph 1.F.3, the Project Gutenberg Literary Archive
Foundation, the owner of the Project Gutenberg™ trademark,
and any other party distributing a Project Gutenberg™ electronic
work under this agreement, disclaim all liability to you for
damages, costs and expenses, including legal fees. YOU AGREE
THAT YOU HAVE NO REMEDIES FOR NEGLIGENCE, STRICT
LIABILITY, BREACH OF WARRANTY OR BREACH OF CONTRACT
EXCEPT THOSE PROVIDED IN PARAGRAPH 1.F.3. YOU AGREE
THAT THE FOUNDATION, THE TRADEMARK OWNER, AND ANY
DISTRIBUTOR UNDER THIS AGREEMENT WILL NOT BE LIABLE
TO YOU FOR ACTUAL, DIRECT, INDIRECT, CONSEQUENTIAL,
PUNITIVE OR INCIDENTAL DAMAGES EVEN IF YOU GIVE
NOTICE OF THE POSSIBILITY OF SUCH DAMAGE.

1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If you

discover a defect in this electronic work within 90 days of
receiving it, you can receive a refund of the money (if any) you
paid for it by sending a written explanation to the person you
received the work from. If you received the work on a physical
medium, you must return the medium with your written
explanation. The person or entity that provided you with the
defective work may elect to provide a replacement copy in lieu
of a refund. If you received the work electronically, the person
or entity providing it to you may choose to give you a second
opportunity to receive the work electronically in lieu of a refund.
If the second copy is also defective, you may demand a refund
in writing without further opportunities to fix the problem.

1.F.4. Except for the limited right of replacement or refund set

forth in paragraph 1.F.3, this work is provided to you ‘AS-IS’,
WITH NO OTHER WARRANTIES OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.

1.F.5. Some states do not allow disclaimers of certain implied

warranties or the exclusion or limitation of certain types of
damages. If any disclaimer or limitation set forth in this
agreement violates the law of the state applicable to this
agreement, the agreement shall be interpreted to make the
maximum disclaimer or limitation permitted by the applicable
state law. The invalidity or unenforceability of any provision of
this agreement shall not void the remaining provisions.

1.F.6. INDEMNITY - You agree to indemnify and hold the

Foundation, the trademark owner, any agent or employee of the
Foundation, anyone providing copies of Project Gutenberg™
electronic works in accordance with this agreement, and any
volunteers associated with the production, promotion and
distribution of Project Gutenberg™ electronic works, harmless
from all liability, costs and expenses, including legal fees, that
arise directly or indirectly from any of the following which you
do or cause to occur: (a) distribution of this or any Project
Gutenberg™ work, (b) alteration, modification, or additions or
deletions to any Project Gutenberg™ work, and (c) any Defect
you cause.

Section 2. Information about the Mission

of Project Gutenberg™
Project Gutenberg™ is synonymous with the free distribution of
electronic works in formats readable by the widest variety of
computers including obsolete, old, middle-aged and new
computers. It exists because of the efforts of hundreds of
volunteers and donations from people in all walks of life.

Volunteers and financial support to provide volunteers with the

assistance they need are critical to reaching Project
Gutenberg™’s goals and ensuring that the Project Gutenberg™
collection will remain freely available for generations to come. In
2001, the Project Gutenberg Literary Archive Foundation was
created to provide a secure and permanent future for Project
Gutenberg™ and future generations. To learn more about the
Project Gutenberg Literary Archive Foundation and how your
efforts and donations can help, see Sections 3 and 4 and the
Foundation information page at www.gutenberg.org.

Section 3. Information about the Project

Gutenberg Literary Archive Foundation
The Project Gutenberg Literary Archive Foundation is a non-
profit 501(c)(3) educational corporation organized under the
laws of the state of Mississippi and granted tax exempt status
by the Internal Revenue Service. The Foundation’s EIN or
federal tax identification number is 64-6221541. Contributions
to the Project Gutenberg Literary Archive Foundation are tax
deductible to the full extent permitted by U.S. federal laws and
your state’s laws.

The Foundation’s business office is located at 809 North 1500

West, Salt Lake City, UT 84116, (801) 596-1887. Email contact
links and up to date contact information can be found at the
Foundation’s website and official page at
www.gutenberg.org/contact
 Section 4. Information about Donations to
the Project Gutenberg Literary Archive
Foundation
Project Gutenberg™ depends upon and cannot survive without
widespread public support and donations to carry out its mission
of increasing the number of public domain and licensed works
that can be freely distributed in machine-readable form
accessible by the widest array of equipment including outdated
equipment. Many small donations ($1 to $5,000) are particularly
important to maintaining tax exempt status with the IRS.

The Foundation is committed to complying with the laws

regulating charities and charitable donations in all 50 states of
the United States. Compliance requirements are not uniform
and it takes a considerable effort, much paperwork and many
fees to meet and keep up with these requirements. We do not
solicit donations in locations where we have not received written
confirmation of compliance. To SEND DONATIONS or determine
the status of compliance for any particular state visit
www.gutenberg.org/donate.

While we cannot and do not solicit contributions from states

where we have not met the solicitation requirements, we know
of no prohibition against accepting unsolicited donations from
donors in such states who approach us with offers to donate.

International donations are gratefully accepted, but we cannot

make any statements concerning tax treatment of donations
received from outside the United States. U.S. laws alone swamp
our small staff.

Please check the Project Gutenberg web pages for current

donation methods and addresses. Donations are accepted in a
number of other ways including checks, online payments and
credit card donations. To donate, please visit:
www.gutenberg.org/donate.

Section 5. General Information About

Project Gutenberg™ electronic works
Professor Michael S. Hart was the originator of the Project
Gutenberg™ concept of a library of electronic works that could
be freely shared with anyone. For forty years, he produced and
distributed Project Gutenberg™ eBooks with only a loose
network of volunteer support.

Project Gutenberg™ eBooks are often created from several

printed editions, all of which are confirmed as not protected by
copyright in the U.S. unless a copyright notice is included. Thus,
we do not necessarily keep eBooks in compliance with any
particular paper edition.

Most people start at our website which has the main PG search
facility: www.gutenberg.org.

This website includes information about Project Gutenberg™,

including how to make donations to the Project Gutenberg
Literary Archive Foundation, how to help produce our new
eBooks, and how to subscribe to our email newsletter to hear
about new eBooks.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com