World Scientific Proceedings Series on Computer Engineering and Information Science Computational Intelligence in Business and Economics Proceedings of the MS’10 International Conference 7925 tp.indd 5/24/10 3:53 PM World Scientific Proceedings Series on Computer Engineering and Information Science Series Editor: Da Ruan (Belgian Nuclear Research Centre (SCK•CEN) & Ghent University, Belgium Vol 1: Computational Intelligence in Decision and Control edited by Da Ruan, Javier Montero, Jie Lu, Luis Martínez, Pierre D’hondt and Etienne E Kerre Vol 2: Intelligent Decision Making Systems edited by Koen Vanhoof, Da Ruan, Tianrui Li and Geert Wets Vol Computational Intelligence in Business and Economics edited by Anna Gil-Lafuente and José M Merigó Chelsea - Computational Intelligence in Business.pmd 5/12/2010, 11:34 AM World Scientific Proceedings Series on Computer Engineering and Information Science Computational Intelligence in Business and Economics Proceedings of the MS’10 International Conference Barcelona, Spain 15 –17 July 2010 editors Anna M Gil-Lafuente University of Barcelona, Spain José M Merigó University of Barcelona, Spain World Scientific NEW JERSEY 7925 tp.indd • LONDON • SINGAPORE • BEIJING • SHANGHAI • HONG KONG • TA I P E I • CHENNAI 5/24/10 3:53 PM Published by World Scientific Publishing Co Pte Ltd Toh Tuck Link, Singapore 596224 USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library World Scientific Proceedings Series on Computer Engineering and Information Science — Vol COMPUTATIONAL INTELLIGENCE IN BUSINESS AND ECONOMICS Proceedings of the MS’10 International Conference Copyright © 2010 by World Scientific Publishing Co Pte Ltd All rights reserved This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA In this case permission to photocopy is not required from the publisher ISBN-13 978-981-4324-43-4 ISBN-10 981-4324-43-4 Printed in Singapore Chelsea - Computational Intelligence in Business.pmd 5/11/2010, 11:07 AM v The MS’10 Barcelona International Conference is supported by This page intentionally left blank vii PREFACE The Association for the Advancement of Modelling & Simulation Techniques in Enterprises (AMSE) and the University of Barcelona are pleased to present the main results of the International Conference of Modelling and Simulation in Engineering, Economics and Management, held in Barcelona, 15 – 17 July, 2010, through this Book of Proceedings published in the Book Series “World Scientific Proceedings Series in Computer Engineering and Information Science” MS’10 Barcelona is co-organized by the AMSE Association and the University of Barcelona, Spain It is co-supported by the Spanish Royal Academy of Financial and Economic Sciences and the Spanish Ministry of Science and Innovation It offers a unique opportunity for researchers, professionals and students to present and exchange ideas concerning modelling and simulation and related topics and see how they can be implemented in the real world In this edition of the MS International Conference, we want to give special attention to the emerging area of Computational Intelligence In particular, we want to focus on the implementation of these techniques in the Economic Sciences Thus, the title of this book is “Computational Intelligence in Business and Economics” Computational Intelligence is a very broad research area that includes fuzzy set theory, neural networks, evolutionary computation, probabilistic reasoning and chaotic comuting as particular research topics of this discipline The growing importance of Computational Intelligence in the Economic Sciences is obvious when looking to the complex world we are living in Every year new ideas and products are appearing in the markets making them very flexible and with strong unpredicted fluctuations Therefore, in order to deal with our world in a proper way, we need to use models that are able to assess the imprecision and the uncertainty MS’10 Proceedings is constituted by 88 papers selected from 141 submissions from 36 countries, making an acceptance rate of 62% We have also included a summary of the presentation given by the plenary speakers: Jaime Gil Aluja, Janusz Kacprzyk and Korkmaz Imanov The book is divided in parts: (1) Theoretical Foundations, (2) Accounting and Finance, (3) Management, (4) viii Marketing, Sports and Tourism, (5) Economics and Politics, (6) Applications in Engineering, and (7) Applications in Other Fields We would like to thank all the contributors, referees and the scientific and honorary committees for their kind co-operation with MS’10 Barcelona; to Jaime Gil Aluja for his role as the President of AMSE and the President of the honorary committee; to the whole team of the organizing committee, including Lluis Amiguet, Luciano Barcellos, Carolina Luis Bassa, Josefa Boria, Jaime Gil Lafuente, Mª Carmen Gracia, Aras Keropyan, Pilar López-Jurado, Onofre Martorell, Carles Mulet, Camilo Prado, Mª Luisa Solé and Emilio Vizuete; and to Chelsea Chin (Editor, World Scientific) for her kind advise and help to publish this volume Finally, we would like to express our gratitude to World Scientific and in particular to Da Ruan (editor-in-chief of the book series “WS Proceedings Series in Computer Engineering and Information Science”) for his support in the preparation of this book Anna M Gil Lafuente, MS’10 Barcelona Chair José M Merigó, MS’10 Barcelona Co-chair Barcelona, March 2010 ix HONORARY COMMITTEE Special thanks to all the members of the Honorary Committee for their support in the organization of the MS’10 Barcelona International Conference President of the Honorary Committeee Jaime Gil-Aluja President of AMSE and President of the Spanish Royal Academy of Financial and Economic Sciences Honorary Committee André Azoulay Ernest Benach Alessandro Bianchi José Casajuana Jacques Delruelle Ricardo Díez Hotchleitner Isidre Fainé Casas Lorenzo Gascón Mohamed Laichubi Juan José Pinto Dídac Ramírez Eugen Simon Lotfi A Zadeh Le Conselleir de Sa Majesté le Roi du Royaume du Maroc President of the Parlament de Catalunya Ex Rector of the University of the Mediterranean Studies of Reggio di Calabria President of the Royal Academy of Doctors President of the Censor School of Accounting of Belgium’s National Bank Honorary President of the Club de Roma President of La Caixa Vicepresident of the Spanish Royal Academy of Financial and Economic Sciences Former Minister and Algerian Ambassador Former President of Caixa Barcelona Rector of the University of Barcelona President of the National Foundation for the Science and the Art of Romania University of California at Berkeley -0.2 X, Y y>0 0.2 0.4 x0 y0 y0 -0.4 y0 -0.4 -0.2 0.0002 N=0.8 Re=550 -0.2 X, Y 0.2 0.4 N=10 Re=550 x0 5E-05 -5E-05 5E-05 pressure-coefficient 0.4 pressure-coefficient X, Y 0.2 0.0001 N=0.2 Re=550 x0 -0.0001 y0 -0.2 pre ssure -coe fficient 5E-05 x>0 -5E-05 pressure-coefficient N=0.4 Re550 x0 y0 N=0.05 Re=550 x0 is a small parameter Now, using the constructed transformation of variables: z = a1 , a2 T qɺM + b10 + g10 , b20 + g 20 T qM , κ1 = a1 , a2 , a3 T qɺ M , κ = a4 qɺM , qj=qj, (j = 1, 4) where ai, bi, gi (i = 1,…,4) are submatrices of matrices a, b, g correspondingly, we shall lead equations (3) to the singularly perturbed form This transformation is the non-linear, non-singular under condition that bi0, j + gi0, j j = 2,3 i =1,2 ≠ , evenly regular [6], not changing the statement of the stability problem System (3) in new variables has a form (1) dz = Z (t , µ , z , x ), dt where x = x1 , x2 , x3 T , M (µ ) x1 = κ1 , q1 T dx = P ( µ ) + X (t , µ , z , x ) dt (4) , x2=κ2, x3=q4; α1=0, α2=2, α3=0; P2i ( µ ) = µ P2′i ( µ ) (i=1,2) The characteristic equation has m zero-roots Other roots can be found from the equation d(λ, µ)=0 We assume the shortened system of 0-level (degenerated system) as an approximate one for a system (4), marking it (4′) without writing In old variables it is the system d ∗ dq a qɺ + (b∗ + g ∗ )qɺ + c∗ q = Q∗ , = qɺ dt dt where q = q1 , q2 , q3 T (5) is s-dimensional vector of generalized coordinates, describing the state of an absolutely rigid system; a*, b*, c*, g* are s×s-matrices of absolutely rigid system The equation (5) describes a motion of an idealized model of mechanical system This model corresponds to an approximate system (4′) of 0-level We shall call it a “limit model” A problem: in what conditions a transition from the initial model (3) to its idealized model (to absolutely rigid system) is possible? Using methods of stability theory [1, 2], combined with the singular perturbations methods [7,8] and introducing the differential equations for deviations that respond to non-critical (basic) variables x, we can find out the acceptability conditions for transition validity from system (4) to the system (4′) in concrete dynamical problems After returning to old variables, taking into account the properties of the considered mechanical system, we receive the corresponding statements 800 3.1 Stability problem When the stability property for reduced model (5) will be ensuring same property for original (full) model (3)? Theorem If bi0, j + g i0, j j = 2,3 i =1,2 ≠ 0, c31 ≠ and all roots (except m zero roots) of characteristic equation of reduced system (5) have negative real parts, then with sufficiently small values of µ (sufficiently high rigidity of the system elements) the zero solution stability of the full system will be succeeding from the zero solution stability of reduced system (5) And reduced system (5) has integral a1∗ a2∗ qɺ + b1∗0 + g1∗0 b2∗0 + g 2∗0 q + ϕ ( q , qɺ ) = B and full system (3) has integral of Lyapunov: a1 a2 3.2 qɺ M + b10 + g10 b20 + g 20 qM + F ( qM , qɺ M ) = A Estimations of approximate solutions Let qi = qi (t , µ ), qɺi = qɺi (t , µ ), (i = 4) be the solution of system (3) with the initial conditions ∗ i ∗ i ⋅∗ i ⋅∗ i q = q (t ) , q = q (t ), qi = qi (t0 , µ ), qɺi = qɺi (t0 , µ ) ; we shall designate (i =1,…4) as the solution of approximate system (5), defined by the initial conditions q∗j = q∗j (t0 ), ∗ q ≡ 0, qɺ ∗j = qɺ ∗j (t0 ) (j=1, 2, 3), where ∗ qɺ ≡ Making use of stability theory methods we can prove the following statement: Theorem If the characteristic equation for system (5) has all roots in the left half-plane (except m zero roots) for d(0, 0)≠0, then under sufficiently big stiffness of the system elements (i.e µ is sufficiently small) there exists such a µ∗-value for ξ>0, η>0, γ>0 given in advance (no matter how small ξ and γ are), that in a perturbed motion: qi − qi∗ < ξ , (i=1,…,4) when 0