Vous êtes sur la page 1sur 3

The Kantorovich Inequality Author(s): Vlastimil Ptak Reviewed work(s): Source: The American Mathematical Monthly, Vol.

102, No. 9 (Nov., 1995), pp. 820-821 Published by: Mathematical Association of America Stable URL: http://www.jstor.org/stable/2974512 . Accessed: 24/01/2012 16:05
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Mathematical Association of America is collaborating with JSTOR to digitize, preserve and extend access to The American Mathematical Monthly.

http://www.jstor.org

[P] [V]

M. Postnikov, Leg,ons geometrie: de Groupes algebres Lie, Editions MIR, Moscou, 1982, et de 1985. V. S. Varadarajan, Lie groups,Lie algebras, and theirrepresentations, Prentice-Hall, Englewood Cliffs, 1974; Reprinted as GTM volume 102, Springer Verlag, Berlin.

URAD0751 au CNRS& UFRde Mathematiques Universite Lille I de F-59655 Villeneuve d'AseqCedex France gmt( gat.univ-lillel .fr W

The KantorovichInequality
VlastimilPtak
lwhe inequality appears first in a survey article on functional analysis and applied mathematics by L. V. Kantorovic; it is used in investigations concerning the condition number of operators and has important applications in estimating convergence of methods of steepest descent for solving equations. In a number of subsequent papers the connection of the inequality with an inequality given by Polya and Szego was cleared up and a number of proofs, some of considerable complexit, of the inequality and of different variants thereof appeared in the literature. In view of the importance of the inequality one more note on the subject might be of interest. It is not difficult to see that the result is essentially based on the inequalit,r between the geometric and arithmetic mean; to emphasise this we restate it in a form using the two means which immediately suggests a simple and natural proof. The Kantorovich inequalit. Supposexl < x2 < * t < x,t are givenpositivenumbers.Let A1,..., A,l> O and LAj = 1. Then

( where A-2(X1
+

Ajxj)(E

AjXJ-1) < A2G-2

x,l) and G =

(x1X,l)l/2.

Proof: Observe that the inequalit,r is homogeneous in the sense that it is invariant with nespect to replacing each x; by a positive multiple arx;.Accordingly it is possible to assume that G = 1 so that x,l = l/xl. Each x between xl and l/xl
satisfies
1
X +X <X1 + X1

It follows that LAjxj + LAjx-1 < xl + l/xl = 2A. The conclusion follows by an application of the geometric- arithmetic mean inequality.

820

NOTES

[November

REFERENCES inequality,Numer. Math. 3 (1961), 1. F. L. Bauer: A further generalizationof the Kantorovich 117-119. 2. F. L. Bauer and A. S. Householder:Some inequalitiesinvolvingthe euclidean conditionof a matnx, Numer.Math.2 (1960),308-311. Proc. of 3. W. Greub and W. Reinboldt:On a generalization an inequalityof L. V. Kantorovich, Amer. Math.Soc. 10 (1959),407-413. 68 Amer.Math.MonthEy, (1961),904-906. inequality, on 4. P. Henrici:Two remarks the Kantorovich 5. A S. Householderand F. L. Bauer: On certain iterativemethods for solving linear systems, Numer.Math.2 (1960),55-59. mat. nauk,3 (in Functionalanalysisand appliedmathematics Russian),Uspechi 6. L. V. Kantorovic: (1948),89-185. 64B inequality,J. Res. Natl. Bur. Standards, (1960),33-34. 7. M. Newman:Kantorovich's Berlin,1925. und derAnatysts, 8. G. Polyaand G. Szego: Aufgaben Lehrsatze inequality, Numer.Math.2 (1960),344-346. 9. A H. Schopf:On the Kantorovich inequality,Proc. Amer.Math.Soc. 11 (1960),468. 10. W. G. Strang:On the Kantorovich

Academyof Sciencesof the CzechRepublic Institute Mathematics of Zitna25 11567Praha1 The CzechRepublic

On The GeneralizedInverse Form of the Equations of ConstrainedMotion


RobertKalaba and RongXu
It 1. INTRODUCTION. has often been observedthat seeminglyabstractconcepts The importancein practicalapplications. and principlesprove to be of paramount close ties between constrainedmotion and generalizedinversesof matricesmaybe a case in point. In 1829 C. F. Gauss formulatedhis celebratedprincipleof least constraintfor handling static and dynamicproblems for constrainedmechanicalsystems. The problem.The seeminglyabstractnotion principletakes the form of a minimization of the generalizedinverse of a matrixproves to be crucialin dealingwith Gauss' principle and in understandingthe complex interactionsbetween applied and constraintforces. In this note, we present Gauss'principleand then indicatethe role of generalized inversesin its furtherdevelopment. I7JFDINVERSESOF MATRICES[1]. 2. GAUSS'PRINCIPLEAND GENERAX Considera system of p particles.Let the mass of the ith particlebe mi and the externalforce acting on it be ti. We use Cartesiancoordinates.If there were no

19951

NOTES

821

Vous aimerez peut-être aussi