• Media type: E-Article
  • Title: Conditions for the Uniqueness of the Fixed Point in Kakutani's Theorem
  • Contributor: Schirmer, Helga
  • imprint: Canadian Mathematical Society, 1981
  • Published in: Canadian Mathematical Bulletin
  • Language: English
  • DOI: 10.4153/cmb-1981-053-5
  • ISSN: 1496-4287; 0008-4395
  • Keywords: General Mathematics
  • Origination:
  • Footnote:
  • Description: <jats:title>Abstract</jats:title><jats:p>Kakutani's Theorem states that every point convex and use multifunction <jats:italic>ϕ</jats:italic> defined on a compact and convex set in a Euclidean space has at least one fixed point. Some necessary conditions are given here which <jats:italic>ϕ</jats:italic> must satisfy if <jats:italic>c</jats:italic> is the unique fixed point of <jats:italic>ϕ</jats:italic>. It is e.g. shown that if the width of <jats:italic>ϕ</jats:italic>(<jats:italic>c</jats:italic>) is greater than zero, then <jats:italic>ϕ</jats:italic> cannot be lsc at <jats:italic>c</jats:italic>, and if in addition <jats:italic>c</jats:italic> lies on the boundary of <jats:italic>ϕ</jats:italic>(<jats:italic>c</jats:italic>), then there exists a sequence {x<jats:sub>k</jats:sub>} which converges to <jats:italic>c</jats:italic> and for which the width of the sets <jats:italic>ϕ</jats:italic>(x<jats:sub>k</jats:sub>) converges to zero. If the width of <jats:italic>ϕ</jats:italic>(<jats:italic>c</jats:italic>) is zero, then the width of <jats:italic>ϕ</jats:italic>(x<jats:sub>k</jats:sub>) converges to zero whenever the sequence {x<jats:sub>k</jats:sub>} converges to <jats:italic>c</jats:italic>, but in this case <jats:italic>ϕ</jats:italic> can be lsc at <jats:italic>c</jats:italic>.</jats:p>
  • Access State: Open Access