Title
Author(s)
Citation
Issue Date
URL
Ky Fan's Inequality for Set-Valued Maps with Vector-Valued
Images (Nonlinear Analysis and Convex Analysis)
Georgiev, Pando Gr.; Tanaka, Tamaki
数理解析研究所講究録 (2001), 1187: 143-154
2001-01
http://hdl.handle.net/2433/64679
Right
Type
Textversion
Departmental Bulletin Paper
publisher
Kyoto University
数理解析研究所講究録
1187 巻 2001 年 143-154
143
Ky Fan’s Inequality for Set-Valued Maps with
Vector-Valued Images*
PANDO
$\mathrm{G}\mathrm{R}$
.
$\mathrm{G}\mathrm{E}\mathrm{o}\mathrm{R}\mathrm{G}\mathrm{I}\mathrm{E}\mathrm{v}\dagger_{\mathrm{a}\mathrm{n}\mathrm{d}}$
TAMAKI TANAKA
(ffl
$\zeta \mathrm{F}$
$\ovalbox{\tt\small REJECT}_{)}\ddagger$
Abstract:
We consider four variants of Fan’s type inequality for vector-valued
multifunctions in topological vector spaces with respect to a cone preorder in the
target space, when the functions and the cone possess various kinds of semicontinuity
and convexity properties. In order to establish these results, firstly we prove a twofunction result of Simons directly by the scalar Fan’s inequality, after that, by its help
we derive a new two-function result, which is the base of our proofs. As a consequence
of our Fan’s type inequalities we obtain that this new two-function result is equivalent
to the scalar Fan’s inequality.
Key words:
pings,
Fan’s inequality, vector-valued multifunctions, semicontinuous map-
vex functions.
$\mathrm{q}\mathrm{u}\mathrm{a}\mathrm{s}\mathrm{i}\mathrm{c}\mathrm{o}\dot{\mathrm{n}}$
1.
Introduction.
This paper is concerned with vector-valued variants of the following type of inequality:
if $f(x, x)\leq 0$ for all , then
,
$x$
$\min_{y\in X_{x}}\sup_{\in X}f(x, y)\leq 0$
which is equivalent to the famous Fan’s minimax inequality (this equivalence was proved by
Takahashi [8, Lemma 1] firstly). This inequality is one of the main tools in the nonlinear and
convex analysis, equivalent to Brouwer’s fixed point theorem, Knaster-Kuratowski-Mazurkiewicz
theorem, and so on. As an analytical instrument, in many situations it is more appropriate
and applicable than other main theorems in nonlinear analysis. We refer to [2] for various type
equivalent theorems in nonlinear analysis.
In this paper we show four kinds of vector-valued Fan’s type inequality for multifunctions.
One of them (Theorem 3.1) generalizes the main result of Ansari-Yao in [1], namely, the existence
result in the so-called Generalized Vector Equilibrium Problem. Any of our Theorems 3.1-3.4
implies the classical Fan inequality, while the main result in [1] does not imply it in its full
generality, but only for continuous functions. Our proofs are quite different from that in [1] and
are based on the classical scalar Fan inequality. More precisely, in the proofs we use a new result
(see Theorem 2.3) which follows from a two-function result of Simons [7, Theorem 1.2] (used in
*This work is based on research 11740053 supported by Grant-in-Aid for Scientific Research from the Ministry
of Education, Science, Sports and Culture of Japan. The first named author is supported by JSPS fellowship
and International Grant for Research in 1999 and 2000 at Hirosaki University. He is very grateful for the warm
hospitality of the University, during his stay as a Visiting Professor.
\dagger Department of Mathematics and Informatics, University of Sofia, 5 James Bourchier Blvd., 1126 Sofia,
-mail: pandogg@fmi.
Current E-mail:
-sofia.bg
Bulgaria
georgiev@bsp.brain.riken.go.jp
\ddagger Department of Mathematical System Science, Faculty of Science and Technology, Hirosaki Universg
)
sity, Hirosaki 036-8561, Japan(
-mail: sltana@cc.hirosaki-u.ac.jp
.
; Secondary:
2000 Mathematic Subject Classiflcation. Primary:
$(7^{\grave{\backslash }})\triangleright j\mathrm{j}\mathrm{t})7\cdot\backslash y7\tau^{\nu}7\star 5\ovalbox{\tt\small REJECT}\Phi \mathrm{E}\mathrm{f}\mathrm{f}\mathrm{l}\neq\infty \mathrm{g}\beta)$
$E$
$\mathrm{u}\mathrm{n}\mathrm{i}$
$\overline{\mathrm{T}}036-8561\ovalbox{\tt\small REJECT} \mathrm{f}\mathrm{f}\mathrm{i}\ovalbox{\tt\small REJECT} \mathrm{f}\mathrm{f}\mathrm{i}\mathrm{f}\mathrm{f}\mathrm{l}\hslash \mathrm{X}\mathrm{B}\mathrm{W}3$
$5.\Lambda \mathrm{f}\mathrm{f}\mathrm{i}|\star\not\in \mathrm{E}\mathrm{I}\mp\#\propto\beta$
$E$
$49\mathrm{J}53$
$49\mathrm{J}35,47\mathrm{H}04$
$\sqrt[\backslash ]{}7_{\backslash }\overline{\tau}\Delta\#.\}^{\mathrm{R}}+\#\}$
144
Fig. 1:
$f(x, y^{*})\not\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
.
Fig.2:
$f(x, y^{*})\cap$
int $C(x)=\emptyset$ .
Fig.3:
$f(x, y^{*})\cap(-C(x))\neq\emptyset$
.
Fig.4:
$f(x, y^{*})\subset(-C(x))$
.
[7] to derive Fan’s inequality), which we prove directly by Fan’s inequality. For a simple proof of
the classical Fan inequality, based on Brouwer’s fixed point theorem, we refer to [3] and ??.
Our main tool in this paper (Theorem 2.3) is a slightly more general form of a two-function
result of Simons [7, Corollary 1.6] and as a consequence of our results, it implies the classical Fan
inequality.
The proofs of the main results (Theorems 4.1-4.4) use Theorem 2.3 for special scalar functions possessing semicontinuity and convexity properties, inherited by the semicontinuity and the
be regarded as
convexity properties of the multifunctions. The four types of Fan’s
nonpositivity
of
of the scalar
the
substituting
by
inequality
Fan’s
generalizations of the classical
function $(f(x, y)\leq 0)$ by various types of set relations between the images of multifunction and
cone; see Figures 1-4.
$\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}\cdot \mathrm{c}\mathrm{a}\mathrm{n}$
Fan’s inequality and a new two-function result.
2.
Theorem 2.1 (Fan). Let $X$ be a nonempty compact convex subset of a topological vector space
be quasiconcave in its first variable and lower semicontinuous in its second
and :
variable. Then
$X\cross Xarrow \mathrm{R}$
$f$
.
$\min_{y\in X_{x}}\sup_{\in X}f(x, y)\leq\sup_{x\in X}f(x, x)$
Theorem 2.2 (Simons [7, Theorem 1.2]). Let be a nonempty compact convex subset
lower semicontinuous in its second variable,
topological vector space, :
Z. Then
quasiconcave in its first variable, and $f\leq g$ on
$Z$
$f$
of a
$g:Z\cross Zarrow \mathrm{R}$
$Z\cross Zarrow \mathrm{R}$
$Z\cross$
$\min_{y\in Z_{x}}\sup_{\in Z}f(x, y)\leq\sup_{z\in Z}g(z, z)$
Proof. Define the function co
co
$f$
.
as a quasiconcave envelope of with respect to the first variable:
$f$
$f(x, y):= \sup\{\min_{1i\in\{,\ldots,n\}}f(x_{i}, y):x=\sum_{i=1}^{n}\lambda_{i^{X_{i},X_{i}}}\in Z, \lambda_{i}\geq 0, \sum_{i=1}^{n}\lambda_{i}=1, n\in \mathrm{N}\}$
,
where is the set of the natural numbers. This function satisfies the conditions of Fan’s
and applying the latter, we obtain the result.
Now we prove our main tool in this paper. Its proof is similar to that of [7, Corollary 1.6].
$\mathrm{N}$
$\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}\mathrm{l}$
Theorem 2.3. Let $X$ be a nonempty compact convex subset of a topological vector space, :
$X\cross Xarrow \mathrm{R}$
quasiconvex in its
lower semicontinuous in its second variable, :
variable, and
$x,$ $y\in X$ and $a(x, y)>0\Rightarrow b(y, x)<0$ .
$a$
$b$
$X\cross Xarrow \mathrm{R}$
$se\backslash cond$
$Su\backslash$
ppose that
$\inf_{x\in X}b(x, x)\geq 0$
. Then there exists
$z\in X$
such that
$a(x, z)\leq 0$
for
all
$x\in X$
.
Proof. The proof is straightforward ffom Theorem 2.2 by defining $f(x, y)=1$ if $a(x, y)>0$
and $f(x, y)=0$ otherwise; $g(x, y)=1$ if $b(y, x)<0$ and $g(x, y)=0$ otherwise.
1
145
3.
Definitions and auxiliary results.
Further let $E$ and $Y$ be topological vector spaces and $F,$ : $Earrow 2^{Y}$ two multivalued mappings
and let for every $x\in E,$ $C(x)$ be a closed convex cone with nonempty interior. We introduce
two types of cone-semicontinuity for set-valued mappings, which are regarded as extensions of
the ordinary lower semicontinuity for real-valued functions; see [5].
Denote
(an open base of int $C(x)$ ), where is a neighborhood
of in , and define the function $h(k, x, y)= \inf\{t : y\in tk-C(x)\}$ , Note that $h(k, x, \cdot)$ is
positively homogeneous and subadditive for every fixed $x\in E$ and $k\in$ int $C(x)$ . Moreover, we
use the following notations $h(k, y)= \inf$ { : $y\in$ tk–C}, and $B=C\cap(2S\backslash \overline{S})$ , where $C$ is
a convex closed cone and is a neighborhood of in $Y$ . Note again that
is positively
homogeneous and subadditive for every fixed
.
Firstly, we prove some inherited properties from cone-semicontinuity.
$C$
$B(x)=(\mathrm{i}\mathrm{n}\mathrm{t}C(x))\cap(2S\backslash \overline{S})$
$0$
$S$
$\mathrm{Y}$
$t$
$S$
$h(k, \cdot)$
$0$
$k\in \mathrm{i}\mathrm{n}\mathrm{t}C$
Definition 3.1. Let
. The multifunction $F$ is
-upper semicontinuous at , if for
every $y\in C(\hat{x})\cup(-C(\hat{x}))$ such that
, there exists an open $U\ni x_{0}$ such that
for every $x\in U$ . If is a Banach space, we shall say that $F$ is $(-C)^{c}$ -upper
semicontinuous at , if for any $\in>0$ and $k\in C$ such that $(k+\in B_{Y}-C)\cap F(x_{0})=\emptyset$ , there
exists $\delta>0$ such that $(k+\in B_{Y}-C)\cap F(x)=\emptyset$ for every $x\in B(x_{0}; \delta)$ .
$\hat{x}\in E$
$C(\hat{x})$
$x_{0}$
$F(x_{0})\subset y+\mathrm{i}\mathrm{n}\mathrm{t}C(\hat{x})$
$Y$
$F(x)\subset y+\mathrm{i}\mathrm{n}\mathrm{t}C(\hat{x})$
$x_{0}$
Definition 3.2. Let
. The multifunction $F$ is
-lower semicontinuous at , if for
every open
such that $F(x_{0})\cap V\neq\emptyset$ , there exists an open $U\ni x_{0}$ such that $F(x)\cap(V+$
int
for every $x\in U$ . If
is a Banach space, we shall say that $F$ is
-lower
$serr\iota icontinuous$ at
, if for any $\in>0$ and $y_{0}\in F(x_{0})$ there exists an open $U\ni x_{0}$ such that
$F(x)\cap(y_{0}+\in B_{Y}+C(\hat{x}))\neq\emptyset$ for every $x\in U$ , where
denotes the open unit ball in Y.
$\hat{x}\in E$
$C(\hat{x})$
$x_{0}$
$V$
$Y$
$C(\hat{x}))\neq\emptyset$
$C(\hat{x})$
$x_{0}$
$B_{Y}$
Remark 3.1. In the two definitions above, the corresponding notions for single-valued function
are equivalent to the ordinary one of lower semicoIltinuity for real-valued function whenever
and $C=[0, \infty)$ . When the cone
consists only of the zero of the space, the notion
in Definition 3.2 coincides with that of lower semicontinuous set-valued mapping. Moreover, it
is equivalent to the cone-lower semicontinuity defined in [5], based on the fact of
$V+C(\hat{x})$ ; see [9, Theorem 2.2].
$Y=\mathrm{R}$
$C(\hat{x})$
$V+\mathrm{i}\mathrm{n}\mathrm{t}C(\hat{x})=$
Proposition 3.1 If for some $x_{0}\in E,$
is a compact subset and multivalued mapping
has a closed graph, then there exists an open set $U\ni x_{0}$ such that
every $x\in U$ . In particular is lower semicontinuous.
$A\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x_{0})$
$W(\cdot):=Y\backslash \{\mathrm{i}\mathrm{n}\mathrm{t}C(\cdot)\}$
$C$
$A\subset C(x)fo\mathit{7}^{\cdot}$
Proof. Assume the coIltrary. Then there exists a net
such that for every
converging to
there exists
. Since
is compact, we may assume that $a_{i}arrow a\in A$ . Since $W$ has
a closed
, it follows that $a\in W(x_{0})$ , which is a contradiction.
$\{x_{i}\}$
$i$
$a_{i}\in A\backslash C(x_{i})$
$x_{0}$
$A$
1
$\mathrm{g}\mathrm{T}\mathrm{a}\mathrm{p}\mathrm{h}$
Lemma 3.1. Suppose that rnultifunction $W$ : $Earrow 2^{Y}$ defined as
closed graph. If the rnultifunction $F$ is $(-C(x))$ -upper semicontinuous at
(the restriction of
the function
$W(x)=Y\backslash \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
$x$
for
each
$x\in E$
has a
, then
$\varphi_{1}|x$
$\varphi_{1}(x).--\inf_{k\in B(x)_{y\in F^{1}(x)}}\mathrm{s}^{\backslash }\mathrm{u}\mathrm{p}h(k, x, y)$
to the set
$(P)$
) is upper
for every
$Y$
If the
$X$
$x\in X$
, that is,
rnapping
$sernicontir\iota uous$
,
if $(F, X)$ satisfies the properiy
there exisis an open
is compact.
$U\ni x$
such that the set
is constant-valued, then
$\varphi_{1}$
;
$F(U\cap X)$
$\overline{F(U\cap X)}$
$C$
$(P)$
is upper semicontinuous.
is precompact in
146
Proof. Assume that $(F, X)$ has property
of
there exists $k_{0}\in B(x_{0})$ such that
$(P)$
. Let
$\in>0$
and
$x_{0}\in X$
be given. By the definition
$\varphi_{1}$
$h(k_{0}, x_{0}, y)<\varphi_{1}(x_{0})+\in$
$\sup$
.
$y\in F^{\gamma}(x_{0})$
Since
$\sup_{y\in F^{\urcorner}(x_{0})}h(k_{0}, x_{0}, y)=\inf\{t:F(x_{0})\subset tk_{0}-C(x_{0})\}$
, we can take
.
$\inf\{t:F(x_{0})\subseteq tk_{0}-C(x_{0})\}<t_{0}<\varphi_{1}(x_{0})+\in$
Since
$F$
is
$(-C(x_{0}))$
-upper semicontinuous at
$F(x)\subset t_{0}k_{0}$
By Proposition 3.1 and property
that
$(P)$
, there exists an open
–int $C(x_{0})$ for every
, for
$t_{0}<t’<\varphi_{1}(x_{0})+\in$
and
$F(x)\subset t’k_{0}-intC(x)$
$x_{0}$
$x\in U$
such that
.
, there exists an open
for every
$k_{0}\in B(x)$
$U\ni x_{0}$
$x\in U_{1}\cap X$
$U_{1}\subset U$
such
.
Then
inf
$=$
$\varphi_{1}.(x)$
$\sup h(k, x, y)$
$k\in B(x)_{y\in F^{7}(x)}$
$h(k_{0}, x, y)$
$\sup$
$\leq$
$y\in t’k0-C(x)$
$=$
$t’h(k_{0}, x, k_{0})+$
$h(k_{0}, x, y)$
$\sup$
$y\in-C(x)$
$\leq$
$t’$
$\leq$
$\varphi_{1}(x_{0})+\in$
.
The proof of the second statement (when is constant-valued) is similar, but in this case there
is no need to use Proposition3.1 and property $(P)$ .
$C$
1
Lemma 3.2. Suppose that the multifunction
and the multifunction $W$ : $Earrow 2^{Y}$ defined by
(the restriction of
function
$F$
is
$-C(x)$ -lower
semicontinuous for each $x\in E$
has a closed graph. Then the
$W(x)=Y\backslash \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
$\varphi_{2}|x$
$\varphi_{2}(x):=\inf_{k\in B(x)}\inf_{y\in F(x)}h(k, x, y)$
to the set $X$ ) is upper semicontinuous, if $(F, X)$ satisfies the property
constant-valued, then
is upper semicontinuous.
$(P)$
.
If the
mapping
$C$
is
$\varphi_{2}$
Proof. Let $\in>0$ and $x_{0}\in E$ be given. By the definition of , for
there exists $k_{0}\in B(x_{0}),$
int $C(x_{0})$ , and $z_{0}\in F(x_{0})$ such that $z_{0}-t_{0}k_{0}\in$ -int $C(x_{0})$ . By
Proposition 3.1, there exists an open set
such that
$\varphi_{2}$
$t_{0}\in(\varphi_{2}(x_{0}), \varphi_{2}(x_{0})+\in)$
$k_{0}\in$
$U_{1}\ni x_{0}$
$z_{0}-t_{0}k_{0}\in$
-int $C(x)$
and
$k_{0}\in \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
for every
$x\in U_{1}$
.
Therefore
$h(k_{0}, x, z_{0})\leq t_{0}$
$U_{2}$
Let $\gamma<\in/2$ . By
such that
$(-C(x_{0}))$
for every
-lower semicontinuity of
$G(x).–F(x)\cap$
[
$z_{0}+\gamma k_{0}$
–int $C(x_{0})$ ]
$F$
$x\in U_{1}$
, there exists an open set
$\neq\emptyset$
for every
Hence
$G(U_{2}\cap X)\subset z_{0}+\gamma k_{0}$
(3.1)
.
–int $C(x_{0})$
$x\in U_{2}$
.
$U_{2}\subset U_{1},$ $x_{0}\in$
(3.2)
147
and
$\overline{G(U_{2}\cap X)}\subset z_{0}+2\gamma k_{0}$
By Proposition 3.1 there exists an open
$U_{3}\subset U_{2},$
$\overline{G(U_{2}\cap X)}\subset z_{0}+2\gamma k_{0}$
–int $C(x_{0})$ .
$U_{3}\ni x_{0}$
such that
–int $C(x)$ for every
$x\in U_{3}$
.
This implies
$F(x)\cap$
Take
$c_{x}\in$
(
$z_{0}+2\gamma k_{0}$
and $y_{x}\in F(x)\cap$ (
-int $C(x)$ . We obtain
$x\in U_{3}\cap X$
$\varphi_{2}(x_{0})+\in$
–int $C(x)$ )
$z_{0}+2\gamma k_{0}$
for every
$\neq\emptyset$
$x\in U_{3}\cap X$
–int $C(x)$ ). Therefore
$\geq$
$t_{0}$
$\geq$
$h(k_{0}, x, z_{0})$
$=$
$h(k_{0}, x, y-2\gamma k_{0}-c_{x})$
$\geq$
$h(k_{0}, x, y)-h(k_{0}, x, 2\gamma k_{0})-h(k_{0}, x, c_{x})$
$\geq$
$h(k_{0}, x, y)-2\gamma$
$\geq$
$\varphi_{2}(x)-\in$
.
$y_{x}=z_{0}+2\gamma k_{0}+c_{x}$
, where
(by (3.1))
(by subadditivity of
$h(k_{0},$ $x,$
$\cdot)$
)
.
Hence
for every
$\varphi_{2}(x_{0})+2\in\geq\varphi_{2}(x)$
$x\in U_{3}\cap X$
.
The proof of the second statement (when $C$ is constant-valued) is similar, but in this case there
is no need to use Proposition 3.1 and property $(P)$ .
1
Lemma 3.3. Suppose that
is a Banach space and the multifunction $F$ : $Earrow 2^{Y}$ is $(-C)^{c}-$
upper semicontinuous and locally bounded (it means that for every point $x_{0}\in E$ there exisis an
open set $U\ni x_{0}$ and $p>0$ such thai $F(x)\subset pB_{Y}$ for every $x\in U$ , where
denotes the
open unit ball in ). Suppose that the multifunction
has a closed graph and the cone $C(x)$
has a compact base $B(x)=(2\overline{B_{Y}}\backslash B_{Y})\cap C(x)$ for every . Then the function
is lower
semicontinuous.
$Y$
$B_{Y}$
$Y$
$C$
$x$
Proof. Firstly we shall prove that the function
tinuous. It is easy to see that
$\varphi_{2}$
$g(k, x):= \inf_{y\in F(x)}h(k, x, y)$
is lower semicon-
$g(k, x)= \inf\{t:(tk-C(x))\cap F(x)\neq\emptyset\}$
for every , we put $g(k,$ $x)=+\infty$ ). Take $(k_{0}, x_{0})\in Y\cross E$ and let
be sequences such that
. Let $\lim$ inf $h(k_{i}, x_{i})=l$ . There exists a
and
subsequence
of
such that $k_{i_{n}}arrow k_{0}\in B(x_{0})$ and $l= \lim g(k_{i_{n}}, x_{i_{n}})$ . Assume
that $l<g(k_{0}, x_{0})$ . Then there exists $\in>0$ such that
(if
$(tk-C(x))\cap F(x)=\emptyset$
$t$
$\{x_{i}\},$ $\{k_{i}\}$
$x_{i}arrow x_{0}$
$\{(k_{i_{n}}, x_{i_{n}})\}$
$k_{i}arrow k_{0}$
$\{(k_{i}, x_{i})\}$
$l+\in<g(k_{0}, x_{0})-\in$
.
(3.3)
By the definition of , there exists
$g$
$y_{i}\in F(x_{i})\cap[(g(k_{i}, x_{i})+\in)k_{i}-C(x_{i})]$
$\forall i\in \mathrm{N}$
.
Hence
(3.4)
$y_{i}=[g(k_{i}, x_{i})+\in]k_{i}-c_{i}$
for some $c_{i}\in C(x_{i})$ . By the locally boundedness of $F$ and from the compactness of $B(x_{0})$ , we
obtain that the sequence
is precompact. Then by (3.4), passing to limits and using the fact
that has a closed graph, we obtain
$\{c_{i}\}$
$C$
$\lim y_{i}=y_{0}=(l+\in)k_{0}-c_{0}$
,
(3.5)
148
is bounded and $B(x_{0})$ is compact, the distance between the sets
and $[g(k_{0}, x_{0})-\in]k_{0}-C(x_{0})$ is positive, so there exists $\alpha>0$ such that
where
$c_{0}\in C(x_{0})$
$F(x_{0})$
. Since
$F(x_{0})$
$([g(k_{0}, x_{0}) - \in]k_{0}+\alpha B_{Y}-C(x_{0}))\cap F(x_{0})=\emptyset$
By the
$(-C)^{c}$
-upper semicontinuity of
$C$
.
we obtain that for some index
$y_{i}\not\in[g(k_{0}, x_{0})-\in]k_{0}+\alpha B_{Y}-C(x_{0})$
$\forall i>i_{0}$
$i_{0}\in \mathrm{N}$
,
.
Hence passing to limit, by (3.3) we obtain $y_{0}\not\in[l+\in]k_{0}-C(x_{0})$ , which is a contradiction with
. Now, we apply Proposition 3.1.21
(3.5). So we proved the lower semicontinuity of at
in [2] and finish the proof.
I
$g$
$(k_{0}, x_{0})$
is a Banach space and the multifunction $F$ : $Earrow 2^{Y}$ is $C(x)-$
Lemma 3.4. Suppose that
lower semicontinuous for each $x\in E$ and locally bounded. Suppose that the multifunction has
a closed graph and the cone $C(x)$ has a compact base $B(x)=(2\overline{B_{Y}}\backslash B_{Y})\cap C(x)$ for every .
Then the function
is lower semicontinuous.
$Y$
$C$
$x$
$\varphi_{1}$
Proof. Firstly we shall prove that the function $g(k, x):= \sup_{y\in
be sequences such that
tinuous. Take
and let
be given. There exists $y_{0}\in F(x_{0})$ such that
$(k_{0}, x_{0})$
$\{x_{i}\},$ $\{k_{i}\}$
$x_{i}arrow x_{0}$
$h(k_{0}, x_{0}, y_{0})>g(k_{0}, x_{0})-\in$
Since
$F$
is
$C$
-lower semicontinuous, for
$\beta>0$
$y_{i}\in F(x_{i})\cap[y_{0}+\beta B_{Y}+C(x_{0})]$
and $b\in B_{Y}$ . Since
, and hence
$c_{i}\in C(x_{0})$
$\in]k_{i}-C(x_{i})$
(3.6)
$i_{0}$
such that
$\forall i>i_{0}$
.
. Hence
$y_{i}=y_{0}+\beta b+c_{i}$
where
and
is lower semicon. Let $\in>0$
$k_{i}arrow k_{0}$
.
there exists index
$F(x_{i})\cap[y_{0}+\beta B_{Y}+C(x_{0})]\neq\emptyset$
Take
F(x)}h(k, x, y)$
,
(3.7)
$y_{i}\in[h(k_{i}, x_{i}, y_{i})+\in]k_{i}-C(x_{i})$
$-y_{0}-\beta b-c_{i}+[g(k_{i}, x_{i})+\epsilon]k_{i}\in C(x_{i})$
, we have
$y_{i}\in[g(k_{i}, x_{i})+$
.
(3.8)
By the locally boundedness of $F$ , from (3.7) and the compactness of $B(x_{0})$ , we obtain that
the sequence
is precompact. Let $\lim$ inf $h(k_{i}, x_{i}, y_{0})=l$ . Without loss of generality (taking
subsequences) we may suppose that $k_{i}arrow k_{0}\in B(x_{0})$ and $l= \lim g(k_{i}, x_{i})$ . Then by (3.8), passing
to limits and using the assumption that has a closed graph, we obtain $y_{0}+\beta b\in(l+\in)k_{0}-C(x_{0})$ .
Hence by (3.6), $g(k_{0}, x_{0})-\in\leq h(k_{0}, x_{0}, y_{0})\leq l+\in+\alpha$ , where $\alpha=h(k_{0}, x_{0}, -\beta b)$ .
), we obtain
are arbitrarily small (therefore is arbitrarily small too, by continuity of
$h(k_{0}, x_{0}, y_{0})\leq l$ .
This proves lower semicontinuity of at
. Now, we apply Proposition
3.1.21 in [2] and finish the proof.
Next, we show some inherited properties from cone-quasiconvexity.
$\{c_{i}\}$
$C$
$\mathrm{S}\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{e}\in>0,$
$h(k_{0},$
$\alpha$
$g$
$x_{0},$
$\beta$
$\cdot)$
$(k_{0}, x_{0})$
1
A multifunction
: $Earrow 2^{Y}$ is called -quasiconvex, if the set { $x\in E$ :
$F(x)\cap(a-C)\neq\emptyset\}$ is convex for every $a\in Y.$
is -quasiconvex, then is said to be
$(-C)$
-quasiconcave, which is equivalent to
-quasiconvex mapping.
Definition 3.3.
$F$
$C$
$\mathrm{I}\mathrm{f}-F$
$C$
$F$
$C$
Remark 3.2. The above definition is exactly that of Ferro type
[6, Definition 3.5].
Definition 3.4. A multifunction
$F:Earrow 2^{Y}$
$(-1)$ -quasiconvex
mapping in
is called (in the sense of [6, Definition 3.6])
149
(a)
$type-(\mathrm{i}\mathrm{i}\mathrm{i})C$
-properly quasiconvex if for every two points
have either
(b)
$F(x_{1})\subset F(\lambda x_{1}+(1-\lambda)x_{2})+C$
or
$x_{2}\in X$
$x_{1},$
and every
$\lambda\in[0,1]$
$F(x_{2})\subset F(\lambda x_{1}+(1-\lambda)x_{2})+C$
we
.
$x_{2}\in X$ and every $\lambda\in[0,1]$ we
-properly quasiconvex if for every two points
have either $F(\lambda x_{1}+(1-\lambda)x_{2})\subset F(x_{1})-C$ or $F(\lambda x_{1}+(1-\lambda)x_{2})\subset F(x_{2})-C$;
$type-(\mathrm{v})C$
$x_{1},$
If $-F$ is type-(iii) [resp. type-(v)] $C$-properly quasiconvex, then $F$ is said be type-(iii) [resp.
type-(v) -properly quasiconcave, which is equivalent to type-(iii) [resp. type-(v)] $(-C)$ -properly
quasiconvex mapping.
$]$
$C$
Remark 3.3. The convexity of (b) above is exactly that of C-quasiconvex-like multifunction in
[1].
Lemma 3.5.
function
If
the
multifunction
$F$
:
is
$Earrow 2^{Y}$
$\psi_{1}(x):=$
inf
$type-(\mathrm{v})C$
-properly quasiconvex, ihen the
$\sup h(k, y)$
$k\in B_{y\in F(x)}$
is quasiconvex.
Proof. By definition, for every
$\lambda)x_{2})\subset F(x_{1})-C$
or
$\lambda\in[0,1]$
and every
$F(\lambda x_{1}+(1-\lambda)x_{2})\subset F(x_{2})-C$
$x_{1},$
$x_{2}\in X$
. Assume that
we have: either
$F(\lambda x_{1}+(1-$
$F(\lambda x_{1}+(1-\lambda)x_{2})\subset F(x_{1})-C$
.
Then
$\psi_{1}(\lambda x_{1}+(1-\lambda)x_{2})$
$:=$
$\inf_{k\in B}\sup\{h(k, y) :
$\leq$
y\in F(\lambda x_{1}+(1-\lambda)x_{2})\}$
$\inf_{k\in B}\sup\{h(k, y) :
inf
$=$
y\in F(x_{1})-C\}$
$h(k, y-c)$
$\sup$
$k\in B_{y\in F()}$
inf
$\leq$
$\sup(h(k, y)+h(k, -c))$
(by subadditivity of
$h(k,$
$\cdot)$
)
$k\in B_{y\in F^{\backslash }(x)}$
$\leq$
$\psi_{1}(x_{1})$
$\leq$
$\max\{\psi_{1}(x_{1}), \psi_{1}(x_{2})\}$
Analogously we proceed in the second case, when
Lemma 3.6.
If
$F$
is
$C$
-quasiconvex, then
.
.
$F(\lambda x_{1}+(1-\lambda)x_{2})\subset F(x_{2})-C$
for every $k\in B$
$\psi_{2}(x;k):=\inf\{h(k, y) :
the
1
function
y\in F(x)\}$
is quasiconvex.
Proof. By the definition of
such that
$\psi_{k}$
, for every
$\in>0$
and
$x_{1},$
$z_{i}-t_{i}k\in-C$
$x_{2}\in E$
there exist
,
$z_{i}\in F(x_{i}),$
$t_{i}\in \mathrm{R}$
(3.9)
and
$t_{i}<\psi_{k}(x_{i})+\in,$
$i=1,2$ .
(3.10)
Since $s_{1}k-C\subset s_{2}k-C$ for
, by (3.9), we have $z_{i} \in t_{i}k-C\subset\max\{t_{1}, t_{2}\}k-C$ . Hence,
by the -quasiconvexity of , for every
there exists $y\in F(\lambda x_{1}+(1-\lambda)x_{2})$ such that
$y \in\max\{t_{1}, t_{2}\}k-C$ , which means
$s_{1}\leq s_{2}$
$C$
$F$
$\lambda\in[0,1]$
$h(k, y)$
$\leq$
$\mathrm{r}\mathrm{n}\mathrm{a}\mathrm{x}\{t_{1}, t_{2}\}$
$<$
$\max\{\psi_{k}(x_{1}), \psi_{k}(x_{2})\}+\in$
150
(by 3.10) and since, the definition, we have
$\psi_{k}(\lambda x_{1}+(1-\lambda)x_{2})=\inf\{h(k, y) :
$\mathrm{a}\mathrm{n}\mathrm{d}\in>0$
is arbitrarily small, we obtain
Lemma 3.7.
function
If the multifunction
$\psi_{2}(x;k)$
or
:
is quasiconcave, where
Proof. By definition, for every
$F(x_{1})+C$
$F$
$\lambda\in[0,1]$
y\in F(\lambda x_{1}+(1-\lambda)x_{2})\})$
I
.
$\psi_{k}(\lambda x_{1}+(1-\lambda)x_{2})\leq\max\{\psi_{k}(x_{1}), \psi_{k}(x_{2})\}$
$Earrow 2^{Y}$
.
$k\in \mathrm{i}\mathrm{n}\mathrm{t}C$
is
$type-(\mathrm{v})C$
-properly quasiconcave, then the
$x_{2}\in X$ we have either
and every
. Assume that $F(\lambda x_{1}+(1-\lambda)x_{2})\subset F(x_{1})+C$ .
$F(\lambda x_{1}+(1-\lambda)x_{2})\subset$
$x_{1},$
$F(\lambda x_{1}+(1-\lambda)x_{2})\subset F(x_{2})+C$
Then
$\psi_{1}(\lambda x_{1}+(1-\lambda)x_{2}; k)$
$=$
$\inf\{h(k, y) :
$\geq$
$\inf\{h(k, y+c) :
$\geq$
$\inf\{h(k, y)-h(k, -c) :
$\geq$
$\inf\{h(k, y) :
$=$
$\psi_{1}(x_{1};k)$
$\geq$
$\min\{\psi_{1}(x_{1};k), \psi_{1}(x_{2)}k)\}$
Analogicaly we proceed in the second case, when
Lemma 3.8.
functio n
If the rnultifunction
$F$
:
y\in F(\lambda x_{1}+(1-\lambda)x_{2})\}$
$k\in \mathrm{i}\mathrm{n}\mathrm{t}C$
y\in F(x_{1}), c\in C\}$
y\in F(x_{1})\}$
.
I
.
$F(\lambda x_{1}+(1-\lambda)x_{2})\subset F(x_{2})+C$
$Earrow 2^{Y}$
is
$type-(\mathrm{i}\mathrm{i}\mathrm{i})C$
$\psi_{1}(x;k):=\sup\{h(k, y) :
is quasiconcave, where
y\in F(x_{1}), c\in C\}$
-properly quasiconcave, then the
y\in F(x)\}$
.
$x_{2}\in X$ we have either
By definition, for every $\lambda\in[0,1]$ and every
$F(x_{2})\subset
F(\lambda
x_{1}+(1-\lambda)x_{2})-C$
or
.
Assume that $F(x_{1})\subset F(\lambda x_{1}+(1-\lambda)x_{2})-C$ . Then
Proof.
$x_{1},$
$F(x_{1})\subset$
$F(\lambda x_{1}+(1-\lambda)x_{2})-C$
$\psi_{2}(x_{1}; k)$
$=$
$\sup\{h(k, y) :
y\in F(x)\}$
$\leq$
$\sup\{h(k, y-c) :
y\in F(\lambda x_{1}+(1-\lambda)x_{2}), c\in C\}$
$\leq$
$\sup\{h(k, y)+h(k, -c) :
$\leq$
$\sup\{h(k, y) :
$=$
$\psi_{2}(\lambda x_{1}+(1-\lambda)x_{2};k)$
y\in F(\lambda x_{1}+(1-\lambda)x_{2}), c\in C\}$
y\in F(\lambda x_{1}+(1-\lambda)x_{2})\}$
,
and hence $\min\{\psi_{2}(x_{1}; k), \psi_{2}(x_{2}; k)\}\leq\psi_{2}(\lambda x_{1}+(1-\lambda)x_{2}; k)$ .
Analogicaly we proceed in the second case, when $F(x_{2})\subset F(\lambda x_{1}+(1-\lambda)x_{2})-C$.
4.
I
Set-valued Fan’s inequalities.
state the main results in this paper. The following theorem is a generalization of that in
[1]. The main difference between our result and that in [1] is the condition (iii), but it allows us
to recover the classical Fan inequality, when $Y$ is the real line. The result in [1] recovers it only
for c\={o}ntinuous functions.
$\mathrm{N}\mathrm{o}\mathrm{w}|\mathrm{w}\mathrm{e}$
Theorem 4.1 Let $K$ be a nonempty convex subset of a topological vector space
logical vector space. Let $F:K\cross Karrow 2^{Y}$ be a multifunction. Assume that
(i)
$C$
:
$Y$
with int
$Karrow 2^{Y}$
is a
muliifunction
$C(x)\neq\emptyset$
;
such that
for every $x\in K,$ $C(x)$
$E,$ $Y$
be a topo-
is a closed convex cone in
151
:
is a
closed in $K\cross Y$ ;
(ii)
$W$
$Karrow 2^{Y}$
multifunction defined as
$W(x)=Y\backslash \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
is $C(x)$ -upper semicontinuous at
for every $y\in K,$
$noi$ constant-valued, then the mapping
the
is
mapping
if
of into precompact subsets of ;
(iii)
$F(\cdot, y)$
$x,$
$C$
$W$
is
with closed values on $K$ and
maps the compact subsets
$Y$
multifunc tion
(iv) there exists a
(b)
of
$F(\cdot, y)$
$K$
(a)
$x$
, and the graph
for
for
every
$x\in K,$
every
$x,$
(c)
$G(x, \cdot)$
(d)
$G(x, y)$
is
$G(x, x)\not\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
$y\in K,$
,
$F(x, y)\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
$type-(\mathrm{v})C(x)$
is compact,
if
such that
$G:K\cross Karrow 2^{Y}$
implies
-properly quasiconcave on
$y\in D$
with
$K$
for
every
,
$x\in X$
,
$G(x, y)\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x)_{i}$
(v) there exists a nonempty compact convex subset
exists
$G(x, y)\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
$F(x, y)\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
.
$D$
of
$K$
such that for every
$x\in K\backslash D$
, there
Then, the solutions set
$S=$
{ $x\in K$ : $F(x,$
is a nonempty and compact subset
$y)\not\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
, for all
$y\in K$
}
of .
$D$
Proof. Put
$h(k, y, z)$ ,
$\sup$
$a(x, y):=- \inf_{k\in By)_{z\in-F(y,x)}}$
$b(x, y):=$
inf
$\sup$
$k\in Bx)_{z\in-G(x,y)}\mathrm{f}$
$h(k, x, z)$ .
It is easy to check that
if and only if
$a(x, y)>0$
by using the compactness of
condition (d), and then $a(x, x)\leq
Denote
, and also
and $b(x, x)\geq
$\overline{F(x,y)}$
0$
$F(y, x)\subset \mathrm{i}\mathrm{n}\mathrm{t}C(y)$
$<0$
$b(y, x)$
0$
if $G(y, x)\subset$ int $C(y)$ by using
.
$S_{y}:=\{x\in D : F(x, y)\not\subset \mathrm{i}\mathrm{n}\mathrm{t}C(x)\}$
(4.1)
.
be a finite
is closed. Let
is lower semicontinuous (by Lemma 3.1), the set
. Obviously $Z$ is compact and convex.
subset of $K$ . Denote by $Z$ the closed convex hull of
Lemmas 3.1, 3.5 and condition (iv) (b) show that the conditions of Theorem 2.3 are satisfied.
Now we apply Theorem 2.3 and obtain a point $z\in Z$ such that $a(y, z)\leq 0$ for every $y\in Z$ ,
which means
(4.2)
for every $y\in Z$ .
Since
$a(y, \cdot)$
$Y_{0}$
$S_{y}$
$Y_{0}\cup D$
$F(z, y)\not\subset \mathrm{i}\mathrm{n}\mathrm{t}C(z)$
The conditions (v) and (4.2) imply that $z\in D$ . Relation (4.1) implies that $\cap\{S_{y} : y\in Y_{0}\}\neq$
. So we proved that the family $\{S_{y} : y\in K\}$ has finite intersection property. Since $D$ is
compact, $\cap\{S_{y} : y\in K\}\neq\emptyset$ , which means that there exists $x_{0}\in K$ such that
int $C(x_{0})$ for every $y\in K$ . So we proved that is nonempty, and since is a closed subset of
$D$ , the proof is completed.
$\emptyset$
$F(x_{0}, y)\not\subset$
$S$
$S$
I
Theorem 4.2. Let $K$ be a nonempty convex subset of a topological vector space
logical vector space, and $F:K\cross Karrow 2^{Y}$ a multifunction. Assume that
(i)
$C$
$Y$
: $Karrow 2^{Y}$ is a
with int
multifunction
$C(x)\neq\emptyset$
;
such that
for
every
$x\in K,$ $C(x)$
$E,$
$Y$
a topo-
is a closed convex cone in
152
is a multifunction defined as
the graph of $W$ is closed in $K\cross Y$ ;
(ii)
$W$
(iii)
for
:
$Karrow 2^{Y}$
$W(x)=Y\backslash \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
,
for every $x\in K$
such that
every $y\in K,$
is $C(x)$ -lower semicontinuous with closed values on $K$ and if
, for every $y\in K$ , maps the
the mapping is not constant-valued, then the mapping
$K$
compact subsets of
into precompact subsets of ;
$F(\cdot, y)$
$x,$
$C$
$F(\cdot, y)$
$Y$
multifunction $G:K\cross Karrow 2^{Y}$
(iv) there exists a
(a)
(b)
(c)
for every $x\in K,$ $G(x, x)\cap$ int
for every $y\in K,$ $F(x, y)\cap$ int
$C(x)=\emptyset$
$x,$
$G(x, \cdot)$
is
on
$C(x)$ -quasiconcave
such that
,
implies
$C(x)\neq\emptyset$
$K$
$G(x, y)\cap$
int
$C(x)\neq\emptyset$
,
for every $x\in K$ ;
(v) there exists a nonempty compact convex subset
.
exists $y\in D$ with
$D$
of
such ihat for every
$K$
, there
$x\in K\backslash D$
$F(x, y)\cap \mathrm{i}\mathrm{n}\mathrm{t}C(x)\neq\emptyset$
Then, the solutions set
$S=$
{ $x\in K$ :
is a nonempty and compact subset
$F(x,$
$y)\cap(\mathrm{i}\mathrm{n}\mathrm{t}C(x))=\emptyset$
, for all
$y\in K$
}
of .
$D$
Proof. Put
$a(x, y):=- \inf_{k\in By)}\inf_{z\in-Fy,x)}h(k, y, z)$
,
$b(x, y):= \inf_{z\in-Gx,y)}h(k(x), x, z)$
where the function is any fixed selection of the multivalued mapping
int $C(x)$ for every $x\in K$ . It is easy to check that
$k$
$a(x, y)>0$
if and only if
$b(y, x)<0$ if and only if
$a(x, x)\leq 0$
,
$xrightarrow \mathrm{i}\mathrm{n}\mathrm{t}C(x)$
$F(y, x)\cap(\mathrm{i}\mathrm{n}\mathrm{t}C(y))\neq\emptyset$
$G(y, x)\cap(\mathrm{i}\mathrm{n}\mathrm{t}C(y))\neq\emptyset$
$b(x, x)\geq 0$
,
, i.e.,
$k(x)\in$
,
,
.
Lemmas 3.2, 3.6 and condition (iv) $(b)$ show that the conditions of Theorem 2.3 are satisfied.
Further the proof is the same as that of Theorem 4.1, but in this case $S_{y}:=\{x\in D$ :
(int $C(x)$ )
.
$F(x, y)\cap \mathrm{I}$
$=\emptyset\}$
Theorem 4.3. Let $K$ be a nonempty convex subset of a topological vector space
space, and $F:K\cross Karrow 2^{Y}$ a multifunction. Assume that
(i)
is a multifunction with a closed graph and
a compact base $B(x)=(2\overline{B_{Y}}\backslash B_{Y})\cap C(x)$ for every ;
$C:Karrow 2^{Y}$
’
$C(x)$
$E,$ $Y$
a Banach
is a closed convex cone with
$x$
(ii)
for every $y\in K,$
(iii) there xists a
$\mathrm{e}$
(a)
(b)
(c)
for
for
$F(\cdot, y)$
is
$(-C)^{c}$
-upper semicontinuous and locally bounded;
multifunction $G:K\cross Karrow 2^{Y}$
every
$x\in K,$ $G(x, x)\cap(-C(x))\neq\emptyset$
every
$x,$
$G(x, \cdot)$
is
such that
,
$y\in K,$ $F(x, y)\cap(-C(x))=\emptyset$
$type-(\mathrm{v})C(x)$
implies
-properly quasiconcave on
(iv) there exists a nonempty compact convex subset
exists $y\in D$ with $F(x, y)\cap(-C(x))=\emptyset$ .
$D$
of
$K$
$G(x, y)\cap(-C(x))=\emptyset$
$K$
for
,
every $x\in K_{f}$.
such that for every
$x\in K\backslash D$
, there
153
Then, the solutions set
{ $x\in K$ : $F(x,$
$S=$
is a nonempty and compact subset
of
$D$
$y)\cap(-C(x))\neq\emptyset$
, for all
$y\in K$
}
.
Proof. Put
inf
$a(x, y):=$
inf
$h(k, y, z)$ ,
inf
$b(x, y):=-$
$k\in B(y)z\in F(y,x)$
$h(k, x, z)$ .
inf
$k\in B(x)z\in G(x,y)$
It is easy to check that
if and only if $F(y, x)\cap(-C(y))\neq\emptyset$ ,
$a(x, y)\leq 0$
$b(y, x)\geq 0$
if and only if
$a(x, x)\leq 0$
$G(y, x)\cap(-C(y))\neq\emptyset$
and
$b(x, x)\geq 0$
,
.
Lemmas 3.3, 3.7 and condition (iii) (b) show that the conditions of Theorem 2.3 are satisfied.
Denote $S_{y}:=\{x\in D : F(x, y)\cap(-C(x))\neq\emptyset\}$ . Since
is lower semicontinuous (by
Lemma 3.3), the set
is closed. Let be a finite subset of $K$ . Denote by $Z$ the intersection of
$K$ and the linear hull of $Y\cup D$ . Obviously
is compact and convex. Now we apply Theorem 2.3
$z\in
Z$
and obtain a point
such that
$a(y, \cdot)$
$Y$
$S_{y}$
$Z$
for every
$a(y, z)\leq 0$
(4.3)
$y\in Z$
which means
for every
$F(z, y)\cap(-C(x))\neq\emptyset$
$y\in Z$
.
(4.4)
Assumption (iv) and condition (4.4) imply that $z\in D$ , and condition (4.4) implies also
. So the family $\{S_{y} : y\in K\}$ has finite intersection property. Since $D$ is
, which completes the proof.
$\cap\{S_{y}$
:
$y\in Y\}\neq\emptyset$
$\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}$
$\cap\{S_{y} :
y\in K\}\neq\emptyset$
Theorem 4.4. Let $K$ be a nonempty convex subset of a topological vector space
space, and $F:K\cross Karrow 2^{Y}$ a multifunction. Assume that
(i)
is a multifunction with a closed graph such that
with a cornpact base $B(x)=(2\overline{B}_{Y}\backslash B_{Y})\cap C(x)$ for every
$C:Karrow 2^{Y}$
$E,$ $Y$
a Banach
is a closed convex cone
$C(x)$
$x_{i}$
(ii)
for every
$x,$
$y\in K,$
(iii) there exists a
(a)
(b)
(c)
$F(\cdot, y)$
is
$C(x)$
-lower semicontinuous and locally bounded;
rnultifunction $G:K\cross Karrow 2^{Y}$
for every $x\in K,$ $G(x, x)\subset-C(x)$ ,
for every $y\in K,$ $F(x, y)\not\subset-C(x)$
$x,$
$G(x, \cdot)$
is
$type-(\mathrm{i}\mathrm{i}\mathrm{i})C(x)$
such that
irnplies
$G(x, y)\not\subset-C(x)$
-properly quasiconcave on
(iv) there exists a nonernpty compact convex subset
exists $y\in D$ with $F(x, y)\not\subset-C(x)$ .
$D$
of
$K$
$K$
for
every
{ $x\in
is a nonernpty and cornpact subset
K:F(x,$
of .
$D$
$y)\subset-C(x)$
$x\in K$ ;
such that for every
Then, the solutions set
$S=$
,
, for all
$y\in K$
}
$x\in K\backslash D$
, ihere
154
Proof. Put
$a(x, y):= \inf_{k\in B(y)}\sup_{z\in F(y,x)}h(k, y, z)$
It is
$\mathrm{e}a\mathrm{s}\mathrm{y}$
,
$b(x, y)$
$:=- \mathrm{i}\mathrm{r}\mathrm{l}\mathrm{f}\sup_{z\in G(x,y)}h(k, x, z)k\in B(x)$
.
to check that
$a(x, y)\leq 0$
if and only if
$b(y, x)\geq 0$
if and only if $G(y, x)\subset-C(y)$ ,
$a(x, x)\leq 0$
and
$F(y, x)\subset-C(y)$ ,
$b(x, x)\geq 0$ .
Lemrnas 3.4, 3.8 and condition (iii) (b) show that the conditions of Theorem 2.3 are satisfied.
Further the proof is the sarne as that of Theorern 4.3, but in this case $S_{y}:=\{x\in D$ : $F(x, y)\subset$
1
$-C(x)\}$ .
5.
Conclusions.
We have presented four type generalizations of the scalar Fan’s inequality in the following setting:
(i) set-valued maps with vector-valued images instead of scalar functions;
(ii) two-function type instead of single function type;
(iii) parametric ordering structure instead of fixed ordering structure;
(iv) cornplete extensions including the result of [1].
As a corollary from any of Theorems 4.1-4.4, we obtain that Theoreni 2.3 irnplies the scalar Fan
inequality.
References
$,[1]$
Q. H. Ansari and J.-C. Yao (1999). An $exister\iota ce$ result for the generalized vector equilibrium
problem, Appl. Math. Lett., 12, 53 56.
[2] J.-P. Aubin and I. Ekeland (1984). Applied Nonlinear Analysis, Wiley Interscience, New
York.
[3] J.-P.Aubin and H. Frankowska (1990). Set- Valued Analysis, Birkh\"auser, Boston.
[4] P. G. Georgiev and T. Tanaka (2000). Vector-valued set-valued variants
ity, J. Nonlinear and Convex Analysis, 1 (3), 245-254.
of Ky Fan’s inequal-
[5] Y. Kimura, K. Tanaka, and T. Tanaka (1999). On semicontinuity of set-valued maps and
marginal functions, 181 188 in Nonlinear Anlysis and Convex Analysis –Proceedings of the
International Conference (W. Takahashi and T. Tanaka, eds.), World Scientific, Singapore.
[6] D. Kuroiwa, T. Tanaka, and T.X.D. Ha (1997). On cone convexity
linear Analysis, TMA, 30(3), 1487-1496.
of set-valued maps,
Non-
[7] S. Simons (1986). Two-function minimax theorems and variational inequalities for functions
on compact and noncompact sets, with some comments on fixed-point theorems, Proc. Symp.
Pure Math., 45 (2), 377-392.
[8] W. Takahashi (1976). Nonlinear variational inequalities and fixed point theorems, JMSJ, 28
(1), 168-181.
[9] T. Tanaka and D. Kuroiwa (1994). Another observation on conditions assuring int $A+B=$
int $(A+B)$ , Appl. Math. Lett., 7 (1), 19-22.