Minimax principles for convex eigenvalue problems (Q1074847)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Minimax principles for convex eigenvalue problems
scientific article

    Statements

    Minimax principles for convex eigenvalue problems (English)
    0 references
    0 references
    1984
    0 references
    Let A be a compact, isotone operator on the positive cone K of a partially ordered locally convex topological vector space E satisfying \(A0>0\). The following problem is considered in the work: To find all numbers \(\lambda >0\) and all \(u\in K\) satisfying \(u=\lambda Au.\) Let \(K^*\) be the dual cone of K and \(\sigma^*=\inf \{1/\lambda:\exists u\in K,u=\lambda Au\}\). It is proved that \(\sigma^*=\inf_{u\in K}\sup_{\phi \in K^*}\phi (Au)/\phi (u)\), and under an additional condition on A (A is convex and maps K into the interior of K) it follows \(\sigma^*=\sup_{\phi \in K^*}\inf_{u\in K}\phi (Au)/\phi (u)\).
    0 references
    0 references
    convex eigenvalue problems
    0 references
    minimax principle
    0 references
    saddle point
    0 references
    characterization
    0 references
    compact, isotone operator
    0 references
    partially ordered locally convex topological vector space
    0 references