Abstract
A Principal Component Analysis is defined for finite metric spaces of L 1-type. For different L 1-criteria, the existence of a solution is proved and a combinatorial algorithm based upon linear programming and graph searching techniques yields such a solution. The efficiency of the algorithm strongly depends not only on the size of the data but also on their nature and on the chosen criterion.
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
AL AYOUBI, B. (1991): Analyse des données en distance de type M 1 . Théorie et algorithmes d’optimisation. Thèse de Doctorat. Université de Haute- Bretagne. Rennes.
ARABIE, Ph. (1991): Was Euclid an unnecessarily sophisticated psychologist? Psychometrika, 56, 4, 567–587.
ASSOUAD, P. & DEZA, M. (1982): Metric subspaces of I1. Publications Mathématiques d’Orsay. Dept. Mathématique. Université de Paris-Sud.
AVIS, D. (1981): Hypermetric spaces and the Hamming Cone. Canadian Journal Math., 33, 4, 795–802.
BANDELT, H. & DRESS, W. (1990): A canonical decomposition theory for metrics on a finite set. Preprint 90–032. Sonderforschungbereich Diskrete Strukturen in der Mathematik. Universität Bielefeld.
BENAYADE, M., FICHET, B. (1993): Dichotomy decompositions of a semi-distance of L 1-type. Rapport de Recherche (to appear). Laboratoire de Biomathématiques. Université d’Aix-Marseille II.
DEZA, M., LAURENT M. (1991): The cut cone: simplicial faces and variety of realizations. Report 9067-OR. Forschungsinstitut für Diskrete Mathematik. Universität Bonn.
EISLER, H. (1973): The algebraic and statistical tractability of the city-block metric. British Journal of Mathematical and Statistical Psychology, 26, 212–218.
FICHET, B. (1987): The role played by L 1 in data analysis. In Y. Dodge ed. Statistical Data Analysis Based on the L 1 -norm and Related Methods. North- Holland, 185–193.
FICHET, B. (1988): L p -spaces in data analysis. In H.H. Bock ed. Classification and Related Methods of Data Analysis. Elsevier Science Publishers, B.V., North- Holland, 439–444.
FICHET, B. (1992): The notion of sphericity for finite L 1-figures of data analysis. In Y. Dodge ed. L 1 -Statistical Analysis and Related Methods. North-Holland, 129–144.
HADELEY, G. (1962): Linear Programming. Addison-Wesley Publishing Company Inc.
HEISER, W. (1987): Correspondence analysis with least absolute residuals. Computational Statistics and Data Analysis, 5, 337–356.
HEISER, W. (1988): Multidimensional scaling with least absolute residuals. In H.H. Bock ed. Classiňcation and Related Methods of Data Analysis. Elsevier Science Publishers, B.V., North-Holland, 455–462.
HEISER, W. (1989): The city-block model for three-way multidimensional scaling. In R. Coppi & S. Bolasco ed. Multiway Data Analysis. North-Holland, 395–404.
HUBERT, L. & ARABIE, Ph. (1986): Unidimensional scaling and combinatorial optimization. In J. De Leeuw & al. ed. Multidimensional Data Analysis. DSWO Press, Leiden, 181–196.
HUBERT, L. & ARABIE, Ph. (1988): Relying on necessary conditions for optimization: unidimensional scaling and some extensions. In H.H. Bock ed. Classification and Related Methods of Data Analysis. Elsevier Science Publishers, B.V., North-Holland, 463–472.
HUBERT, L. & ARABIE, Ph., HESSON-MCINNIS, M. (1992): Multidimensional scaling in the city-block metric: a combinatorial approach. Journal of classification, 9, 2, 211–236.
LE CALVE, G. (1987): L 1-embeddings of a data structure (I, D). In Y. Dodge ed. Statistical Data Analysis Based on the L 1 -norm and Related Methods. North- Holland, 195–202.
MATHAR, R. (1993): Multidimensional scaling with l p-distance matrices. 17 th. Annual Conference of the Gesellschaft für Klassifikation e.V. Kaiserlautern. Germany.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1994 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Benayade, M., Fichet, B. (1994). Algorithms for a geometrical P.C.A. with the L 1-norm. In: Diday, E., Lechevallier, Y., Schader, M., Bertrand, P., Burtschy, B. (eds) New Approaches in Classification and Data Analysis. Studies in Classification, Data Analysis, and Knowledge Organization. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-51175-2_7
Download citation
DOI: https://doi.org/10.1007/978-3-642-51175-2_7
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-58425-4
Online ISBN: 978-3-642-51175-2
eBook Packages: Springer Book Archive