Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • 44 views

    National Academy of Sciences

    Uploaded by

    Aman Machra
    This document presents an alternative proof of a theorem due to John von Neumann regarding the existence and uniqueness of solutions to systems of linear inequalities and equations. The proof reduces the problem to consideration of a single system of linear equations. It does this in two steps: 1) It reframes the systems of inequalities and equations as equivalent adjoint systems of linear equations and inequalities. 2) It utilizes a known relationship between adjoint systems that the first system admits a solution if and only if the second system admits no positive solution. The author argues this alternative proof suggests the theorem could potentially be generalized to replace the finite matrix with a function and summations with more general linear operators.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd

    National Academy of Sciences

    Uploaded by

    Aman Machra
    44 views4 pages
    This document presents an alternative proof of a theorem due to John von Neumann regarding the existence and uniqueness of solutions to systems of linear inequalities and equations. The proof reduces the problem to consideration of a single system of linear equations. It does this in two steps: 1) It reframes the systems of inequalities and equations as equivalent adjoint systems of linear equations and inequalities. 2) It utilizes a known relationship between adjoint systems that the first system admits a solution if and only if the second system admits no positive solution. The author argues this alternative proof suggests the theorem could potentially be generalized to replace the finite matrix with a function and summations with more general linear operators.

    Original Description:

    anything

    Original Title

    87659

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document presents an alternative proof of a theorem due to John von Neumann regarding the existence and uniqueness of solutions to systems of linear inequalities and equations. The proof reduces the problem to consideration of a single system of linear equations. It does this in two steps: 1) It reframes the systems of inequalities and equations as equivalent adjoint systems of linear equations and inequalities. 2) It utilizes a known relationship between adjoint systems that the first system admits a solution if and only if the second system admits no positive solution. The author argues this alternative proof suggests the theorem could potentially be generalized to replace the finite matrix with a function and summations with more general linear operators.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    44 views4 pages

    National Academy of Sciences

    Uploaded by

    Aman Machra
    This document presents an alternative proof of a theorem due to John von Neumann regarding the existence and uniqueness of solutions to systems of linear inequalities and equations. The proof reduces the problem to consideration of a single system of linear equations. It does this in two steps: 1) It reframes the systems of inequalities and equations as equivalent adjoint systems of linear equations and inequalities. 2) It utilizes a known relationship between adjoint systems that the first system admits a solution if and only if the second system admits no positive solution. The author argues this alternative proof suggests the theorem could potentially be generalized to replace the finite matrix with a function and summations with more general linear operators.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 4

    On a Theorem of Von Neumann Author(s): Lloyd L.

    Dines Source: Proceedings of the National Academy of Sciences of the United States of America, Vol. 33, No. 11 (Nov. 15, 1947), pp. 329-331 Published by: National Academy of Sciences Stable URL: http://www.jstor.org/stable/87659 . Accessed: 24/04/2013 15:56
    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.

    National Academy of Sciences is collaborating with JSTOR to digitize, preserve and extend access to Proceedings of the National Academy of Sciences of the United States of America.

    http://www.jstor.org

    This content downloaded from 210.212.58.168 on Wed, 24 Apr 2013 15:56:58 PM All use subject to JSTOR Terms and Conditions

    VOL.33, 1947

    MA THEMATICS: L. L. DINES ON A THEOREM OF VON NEUMANN

    329

    BY LLOYD L. DINES
    DEPARTMENT OF MATHEMATICS, SMITH COLLEGE

    CommunicatedAugust 13, 1947

    In these PROCEEDINGS, August 15, 1946, Loomis' presented a simple of a theorem dud to proof J. von Neumann2' 3 and formulated by Loomis as follows:
    Let aij and bj be two rectangular matrices (i = 1, ..., n), (j = 1, ..., m)

    .such that aij > 0 for all i, j.


    (X,..., xm), y = (y1, ...,

    Then there exists a unique X and vectorsx =


    lm, xj = 1 and Zlyir = 1 O,

    Yn) subject to Xj _ O, y0

    such that m
    X E aijxj
    j=l
    n

    m
    j=l
    n i=1

    E bijxj, i

    = 1, ...,

    n,

    (1)

    X E aijy <
    i=l

    bijyz,j = 1, ..

    w.

    (2)

    The present note gives an alternative simple proof, the essence of which is a reduction of the question to consideration of a single system of linear equations. This reduction is accomplished in two steps: 1st, replacement of (1) and (2) by an equivalent pair of adjoint linear systems
    m

    j=1
    n+m i-1

    ajx O>

    i=

    1, ...,n

    + m,

    (1') (2')

    E Coyi = 0,

    j =1,...,m;

    and 2nd, utilization of a known relationship4 between adjoint systems of form (1'), (2'). This relationship is simply that the system of inequalities (1') admits a solution x which does not annul all the left membersif, and only if, the system of equations (2') admits no positive solution y (everyyi positive). The system (1') consists of n + m inequalities, the first n of which are those in (1) after obvious transpositions, and the last m are the conditions
    xj _ 0, j = 1, ..., m, which occur as collateral conditions in the statement

    of the theorem. The system (2') consists of m equations, which are equivalent to the m inequalities of (2) by virtue of the introduction of m new and essentially
    non-negative variables yn+l, ...,
    yn+m.

    The explicit definitions of the coefficients aio are i = 1, ...;, n; aij = Xaij - bi, = i = n + 1, ..., n + m; ai a,j where bij = 1 if i = n + j; otherwise it is zero.

    1, ..., m, 1, ...,m,

    This content downloaded from 210.212.58.168 on Wed, 24 Apr 2013 15:56:58 PM All use subject to JSTOR Terms and Conditions

    330

    MA THEMA TICS: L. L. DINES

    PROC.N. A. S.

    In view of the relationships which have been indicated above, the von Neumann theorem will be justified if we show the existence of a unique value of the parameter X for which the system of equations (2') admits no positive solution y = (yl, ..., y,+m), but does admit a non-negative
    solution with at least one yk positive.

    To this end, we consider the explicit form of the system (2'), which may be written
    n i=1

    (Xa, -

    bij)yi + yn+j = 0,

    j =

    ...,

    m.

    (2")

    Since aij > 0 for every i, j, it is obvious that for X sufficiently small the system will admit positive solutions y, while for X sufficiently large it will admit no such solution. Furthermore if it admits a positive solution for any particular value of X, it will admit such a solution for any smaller value. We denote by Xothe least upper bound of the values of X for which (2") admits a positive solution y. Since to every positive solution y there
    corresponds by simple normalization a solution for which Zln+myk = 1,

    it follows from the compactness of the closed and bounded sub-space


    defined by yk > 0, Zln+myk = 1 that to the value X =Xo there corresponds a solution of (2") in this subspace; call it y? = (y, , .., y+m).

    At least one of the co6rdinates yk? must be zero. For if they were all positive, a sufficiently small e > 0 would make correspond to the parameter
    value Xo+ e a positive solution y' = (yl', .. ., yn+m') with Yi' = yi?, (i = 1, .., n), and y,,+.' = yn+ j? - e Zlnaijyi0, (i = 1, ..., m), thus contradicting

    the definition of Xo. In Xowe have therefore one value of the parameter X which affords an admissible solution. No second value of X does. For if Xi > Xo did'afford an admissible
    solution y', then 2"l(Xila, - boj)yi' _ O, (j = 1, . .., m), and consequently,

    for any e > 0


    n

    Z
    i=1

    [(x - e)ai - bi]yi'

    < 0

    j = 1, ...,m,

    and from the continuity of the linear functions these inequalities remain
    'strictly valid if each yi' be replaced by yi" = yi' + qi with 71i> 0 and by the equations sufficiently small. Defining y,+ j", (j = 1, ...m)

    i=1

    [(Xi - e)ai -

    bi]yi" + y,+j" = 0

    j = 1,. ..,

    m,

    we would then have a positive solution corresponding to the parameter value X1- e > X0,thus contradicting the definition of X0. This completes the proof of the theorem. The method of proof definitely suggests the possibility of generalization, since the vital relationship between the adjoint systems (1') and (2') persists when the finite matrix

    This content downloaded from 210.212.58.168 on Wed, 24 Apr 2013 15:56:58 PM All use subject to JSTOR Terms and Conditions

    VOL.33, 1947

    MATHEMATICS: L. GARDING

    331

    aij is replaced by a function a(p, q) and the summations are replaced by more general linear operators.
    1Loomis, L. H., "On a Theorem of von Neumann," Proc. Nat. Acad. Sci., 32, 213-215

    (1946). 2 von Neumann, J., "Uber ein okonomisches Gleichungssystem, etc.," Ergebnisse
    eines Mathematischen Kolloquiums, 8, 73-83 (1937). 3 von Neumann, J., and Morgenstern, O., "Theory of Games and Economic Behavior," Princeton University Press (1944).

    4 Dines, L. L., "Convex Extension and Linear Inequalities," Bull. Amer. Math. Soc., 42, 353-365 (1936).

    NOTE ON CONTINUOUS

    REPRESENTATIONS
    BY LARS GARDING

    OF LIE GROUPS

    INSTITUTE

    OF MATHEMATICS,

    LUND,

    SWEDEN,

    AND

    PRINCETON

    UNIVERSITY

    Communicated August 28, 1947

    Let G be an analytic Lie group with unit element e and let a -* T(a), (a e G), be a representation of G as bounded operators T(a) on a Banach space B such that for every x e B, T(e)x = x and T(a)x -> T(b)x when a -> b. Let A = (a(s)), (s real), be a one-parameter subgroup in G and define TAX = lim,_es-l((T(a(s)) )x

    whenever the limit exists in which case x is said to be in the domain of TA. I. Gelfand proved that every such domain is dense in B.1 By a slight extension of his method we will prove that they have a dense intersection when A varies, thus answering a question left open by W. Wigner2 and V. Bargmann.3 Let r be a not-negative integer or + co, let Cr be the class of real functions defined on G with continuous derivatives of order < r, (< r if r = oo), and let Cr?be the subset of Cr whose elements vanish outside a compact set and let Br be the set of elements of B of the form x(h) =
    cG h(b)T(b)xdb

    (h e Cr?, x e B),

    where db is a left invariant volume element on G. THEOREM. Every Br+1 is dense in B, it is in the domain of any TA and TAB,+l is contained in Br for any A. Proof: One has with s = 0 s-'(T(a(s)) l)x(h) = s-l fG h(b)(T(a(s)b) - T(b))xdb = - h(b))T(b)xdb. JG s-l(h(a-(s)b)

    This content downloaded from 210.212.58.168 on Wed, 24 Apr 2013 15:56:58 PM All use subject to JSTOR Terms and Conditions

    You might also like