Abstract
The paper is devoted to the solution of the metric Euclidean Multidimensional Scaling Problem (MDS) and the Euclidean distance geometry problem by methods of d.c. programming developed earlier by the authors. The algorithms, called DCA (DC Algorithms), are based on the duality theory and local optimality conditions for d.c. programming. Different regularization techniques are used in order to improve the qualities (robustness, stability, convergence rate and globality of computed solutions) of DCA. Lagrangian duality without gap allows us to find interesting equivalent forms of (MDS) and a very simple expression of the dual objective function which can be used for checking globality of solutions computed by DCA. DCA (with and without regularization techniques) is described for solving both MDS problems. Requiring only matrix-vector products and only one Cholesky factorization, it is quite simple, and allows exploiting sparsity in large-scale problems. Moreover the well-known majorization method by de Leeuw can be viewed as a special case of DCA. DCA (which globally solves the trust region problem) is also presented in its parametric version to deal with the Euclidean metric MDS problem. Finally extensive numerical experiments are reported which show the robustness and efficiency of DCA for solving the Euclidean metric MDS problem, especially in the largescale setting. They also show the globality of computed solutions in the case where the dissimilarities are the Euclidean distances between the objects.
This work was partially supported by the computer resources financed by Contrat de Plan Interregional du Bassin Parisien-Pole Interregional de Modelisation en Sciences pour Ingenieurs
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
ABDO Y. ALFAKIH, A. KHANDANI & H. WOLKOWICZ, An interior-point method for the Euclidean distance matrix completion problem,Research Report CORR 97–9, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada.
LE THI HOAI AN, Contribution d l’optimisation non convexe et l’optimisation globale. Théorie, Algorithms et Applications, Habilitation à Diriger des Recherches, Université de Rouen, Juin 1997.
LE THI HOAI AN, PHAM DINH TAO and L.D. MUU (1996), Numerical solution for Optimization over the efficient set by D.c. Optimization Algorithm, Operations Research Letters, 19, pp. 117–128.
LE THI HOAI AN and PHAM DINH TAO (1997), Solving a class of linearly constrained indefinite quadratic problems by D.c. algorithms, Journal of Global Optimization, 11 pp. 253–285.
AUSLENDER, Optimisation, méthodes numériques, Ed. Masson. Paris, 1976.
R. BEALS, D.H. KRANTZ & A. TVERSKY, Foundation of multidimensional scaling, Psychol. Rev. Vol 75, pp. 12–142, 1968.
W. BICK, H. BAUER, P.J. MUELLER & O. GIESEKE, Multidimensional scaling and clustering techniques (theory and applications in the social science), Institut fur angewandte sozialforchung, Universitat zu koln, 1977.
H. BREZIS, Opérateurs maximaux monotones, Mathematics Studies 5, North Holland, 1973.
G.M. CRIPPEN, A novel approach to calculation of conformation: distance geometry, J. Computational physics, Vol 24 (1977), pp. 96–107.
G.M. CRIPPEN, Rapid calculation of coordinates from distance measures, J. Computational physics, Vol 26 (1978), pp. 449–452.
G.M. CRIPPEN & T.F. HAVEL, Distance Geometry and Molecular Conformation, Research Studies Press, Taunton, Somerset, UK, 1988.
J. DE LEEUW, Applications of convex analysis to multidimensional scaling, Recent developments in statistics, J.R. Barra et al., editors, North-Holland Publishing Company, pp. 133–145, 1977.
J. DE LEEUW, Convergence of the majorization method for multidimensional scaling, Journal of classification, Vol 5 (1988), pp. 163–180.
R.O. DUDA and P.E. HART, Pattern Classification and Scene Analysis, Wiley, New York, 1973.
W. GLUNT, T.L. HAYDEN, & M. RAYDAN, Molecular conformation from distance matrices, J. Comp. Chem., 14 (1993), pp. 114–120.
L. GUTTMAN, A general nonmetric technique for finding the smallest coordinate space for a configuration of points, Psychometrika, Vol 33 (1968), pp. 469–506.
T.F. HAVEL, An evaluation of computational strategies for use in the determination of protein structure from distance geometry constraints obtained by nuclear magnetic resonance, Prog. Biophys. Mol. Biol., 56 (1991), pp. 43–78.
B. A. HENDRICKSON, The Molecule Problem: Determining Conformation from Pairwise Distances, Ph.D. Thesis, Cornell University, Ithaca, New York, 1991.
B. A. HENDRICKSON, The Molecule Problem: Exploiting structure in global optimization, SIAM J. Optim., 5 (1995), pp. 835–857.
J.B. HIRIART-URRUTY, From convex optimization to non convex optimization. Part I: Necessary and sufficent conditions for global optimality, Nonsmooth Optimization and Related Topics, Ettore Majorana International Sciences, Series 43, Plenum Press, 1988.
R. HORST & H. TUY, Global Optimization (Deterministic Approaches), second edition, Springer-Verlag, Berlin New York, 1993.
H. KONNO, Maximization of a convex quadratic function under linear constraints, Mathematical Programming, 11 (1976), pp. 117–127.
H. KONNO, P.T. THACH and H. TUY, Optimization on Low Rank Nonconvex Structures, Kluwer, Dordrecht-Boston-London, 1997.
J.B. KRUSKAL, Multidimensional scaling by optimizing goodnessof-fit to a non metric hypothesis, Psychometrika, Vol 29, pp. 1–28, 1964.
J.B. KRUSKAL, Nonmetric multidimensional scaling: a numerical method, Psychometrika, Vol 29 (1964), pp. 115–229.
J.B. KRUSKAL & M. WISH, Multidimensional Scaling, Newbury Park, CA. Sage, 1978.
P.J. LAURENT, Approximation et Optimisation, Hermann, Paris, 1972.
M. LAURENT, Cuts, matrix completions and a graph rigidity, Mathematical Programming, Vol. 79 (1997), Nos 1–3, pp. 255–283.
W.J.M. LEVET, J.P. VAN DE GEER & R. PLOMPI, Tridiac comparaisons of musical intervals, Bristish. J. Math. Statist. Psychol. Vol 19 (1966), pp. 163–179.
J.C. LINGOES & E.E. ROSKAM, A mathematical and emprical analysis of two multidimensional scaling algorithms, Psychometrika, Vol 38 (1973), monograph supplement.
P.L. LIONS & B. MERCIER, Splitting algorithms for the sum of two nonlinear operators, SIAM J. Numer. Anal. 16,6 (1979), pp. 964–979.
P. MAHEY and PHAM DINH TAO, Partial regularization of the sum of two maximal monotone operators, Math. Modell. Numer. Anal. (M2AN), Vol. 27 (1993), pp. 375–395.
P. MAHEY and PHAM DINH TAO, Proximal decomposition of the graph of maximal monotone operator, SIAM J. Optim. 5 (1995), pp. 454–468.
B. MIRKIN, Mathematical Classification and Clustering,in Book Series “Nonconvex Optimization arid its Applications”, P.M. Pardalos and R. Horst eds., Kluwer Academic Publishers.
J.J. MORE & ZHIJUN WU, Global continuation for distance geometry problems, SIAM J. Optim., Vol 7, No 3 (1997), pp. 814–836.
J.J. MORE & Z. WU, Issues in large-scale global molecular optimization, preprint MCS-P539–1095, Argonne National Laboratory, Argonne, Illinois 60439.
G.L. NEMHAUSER & L.A. WOLSEY, Integer and Combinatorial Optimization, John Wiley & Sons, 1988.
A.M. OSTROWSKI, Solutions of equations and systems of equations, Academic Press, New York, 1966.
P.M. PARDALOS & J.B. ROSEN, Constrained Global Optimization: Algorithms and applications, Lecture Notes in Computer Science, 268, Springer-Verlag, Berlin, 1987.
PHAM DINH TAO, Eléments homoduaux d’une matrice A relatif à un couple des normes (0,11)). Applications au calcul de SS,f,(A), Séminaire d’analyse numérique, Grenoble, n°236, 1975.
PHAM DINH TAO, Calcul du maximum d’une forme quadratique définie positive sur la boule unité de oc. Séminaire d’analyse numérique, Grenoble, n°247, 1976.
PHAM DINH TAO, Contribution à la théorie de normes et ses applications à l’analyse numérique, Thèse de Doctorat d’Etat Es Science, Université Joseph Fourier- Grenoble, 1981.
PHAM DINH TAO, Convergence of subgradient method for computing the bound norm of matrices, Linear Alg. and Its Appl, Vol 62 (1984), pp. 163–182.
PHAM DINH TAO, Algorithmes de calcul d’une forme quadratique sur la boule unité de la norme maximum, Numer. Math., Vol 45 (1984), pp. 377–401.
PHAM DINH TAO, Algorithms for solving a class of non convex optimization problems. Methods of subgradients, Fermat days 85. Mathematics for Optimization, Elsevier Science Publishers B.V. North-Holland, 1986.
PHAM DINH TAO, Iterative behaviour, Fixed point of a class of monotone operators. Application to non symmetric threshold function, Discrete Mathematics 70 (1988),pp. 85–105.
PHAM DINH TAO, Duality in d.c. (difference of convex functions) optimization. Subgradient methods, Trends in Mathematical Optimization, International Series of Numer Math. Vol 84 (1988), Birkhauser, pp. 277–293.
PHAM DINH TAO et LE THI HOAI AN, Stabilité de la dualité lagrangienne en optimisation d.c. (différence de deux fonctions convexes), C.R. Acad. Paris, t.318, Série I (1994), pp. 379–384.
PHAM DINH TAO and LE THI HOAI AN, Lagrangian stability and global optimality on nonconvex quadratic minimization over Euclidean balls and spheres, Journal of Convex Analysis, 2(1995), pp. 263–276.
PHAM DINH TAO and LE THI HOAI AN (1997), Convex analysis approach to d.c. programming: Theory, Algorithm and Applications (dedicated to Professor Hoang ’Thy on the occasion of his 70th birthday), Acta Mathematica Vietnamica, Vol. 22, N°1, 1997, pp 289–355.
PHAM DINH TAO and LE THI HOAI AN (1998), D.c. optimization algorithms for trust region problem. SIAM J. Optimization, vol 8, No 2, pp. 476–505.
B. POLYAK, Introduction to Optimization. Optimization Software, Inc., Publication Division, New York, 1987.
R.T ROCKAFELLAR, Convex Analysis, Princeton University, Princeton, 1970.
B.D. RIPLEY, Pattern Recognition and Neural Networks, Cambridge University Press, 1996.
R.T ROCKAFELLAR, Monotone operators and the proximal point algorithm,SIAM J. Control and Optimization, Vol.14, N°5 (1976).
J. B. SAXE, Embeddability of Weighted Graphs in k-space is Strongly NP-hard, Proc. 17 Allerton Conference in Communications, Control and Computing, 1979, pp. 480–489.
R.N SHEPARD, Representation of structure in similarity data; problem and prospects, Psychometrika, Vol 39 (1974), pp. 373–421.
J.E. SPINGARN, Partial inverse of a monotone operator, Appl. Math. Optim. 10 (1983), pp. 247–265.
Y. TAKANE, F.W. YOUNG & J. DE LEEUW Nonmetric individual differences multidimensional scaling: an alternating least squares method with optimal scaling features, Psychometrika, Vol 42 (1977), pp. 7–67.
J.F. TOLAND, On subdifferential calculus and duality in nonconvex optimization, Bull. Soc. Math. France, Mémoire 60 (1979), pp. 177–183.
H. TUY, A general deterministic approach to global optimization via d. c. programming, Fermat Days 1985: Mathematics for Optimization, North-Holland, Amsterdam, (1986), pp. 137–162.
H. TUY, Introduction to Global Optimization, Les Cahiers du GERAD, Groupe d’Etudes et de Recherche en Analyse des Décision, Montréal, Québec, 1994.
H. TUY, Convex Analysis and Global Optimization, Kluwer 1998.
R. VARGA Matrix iterative analysis, Prentice Hall, 1962.
A.R. WEBB, Multidimensional scaling by iterative majorization using radial basis functions, Pattern Recognition, 28 (5) (1995), pp. 753–759.
Z. ZOU, RICIIARD.H. BIRD, & ROBERT B. SCHNABEL, A Stochastic/Pertubation Global Optimization Algorithm for Distance Geometry Problems, J. of Global Optimization, 11 (1997), pp. 91–105.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2001 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
An, L.T.H., Tao, P.D. (2001). D.C. Programming Approach to the Multidimensional Scaling Problem. In: Migdalas, A., Pardalos, P.M., Värbrand, P. (eds) From Local to Global Optimization. Nonconvex Optimization and Its Applications, vol 53. Springer, Boston, MA. https://doi.org/10.1007/978-1-4757-5284-7_11
Download citation
DOI: https://doi.org/10.1007/978-1-4757-5284-7_11
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4419-4852-6
Online ISBN: 978-1-4757-5284-7
eBook Packages: Springer Book Archive