Skip to main content

The existence of contingent epiderivative for a set-valued mapping and vector variational-like inequalities

Abstract

In this paper we analyze the existence of the contingent epiderivative for a set-valued mapping in a general normed space with respect to Daniell cone. As its application, we will also establish the relationships between set-valued optimization problems and variational-like inequality problems under the conditions of pseudo invexity.

MSC:49J52, 90C56.

1 Introduction

In [1], Jahn and Raüh introduced the notion of contingent epiderivative for a set-valued mapping, which modifies a notion introduced by Aubin [2] as upper contingent derivative, and also established the existence theory of this contingent epiderivative for a single-valued function. In [3], Jahn and Khan obtained the existence of this kind of contingent epiderivative for a real set-valued function. It has been shown that this notion of contingent epiderivative is a fundamental concept for the formulation of optimality conditions in set-valued optimization, but there are few works that study its existence for a set-valued mapping in general conditions. Although in [4] Rodríguez-Marín and Sama derived the existence of contingent epiderivative, this can only be assured if a set-valued mapping F has the LBD (lower bounded derivative) property. In the last decades, many researchers have given several other generalized notions of epiderivatives by using weak minimizers and minimizers and derived the existence theories for them; see [5] and [4, 6, 7], respectively. Using different kinds of minimal elements, one can define different kinds of epiderivatives. In our paper, we use the ideal minimal elements of a set and the concept of contingent cone to define the contingent epiderivative and analyze its domain, existence, uniqueness and other properties. Under determined conditions, we establish dom(DF( x 0 , y 0 ))=cone(A x 0 )=T(A, x 0 ) and get the existence of DF( x 0 , y 0 ), where domF is the domain of F, DF( x 0 , y 0 ) is the contingent epiderivative of F at ( x 0 , y 0 )grF (the graph of F is denoted by grF) and T(A, x 0 ) is the contingent cone of A at x 0 , respectively.

Our other purpose in this paper is to investigate the relationships between set-valued optimization problems and variational-like inequality problems. In fact, the relationships between vector variational-like inequality problems and optimization problems for a single-valued mapping have been studied by many authors, see [810] and so on; and in [11], Zeng and Li also discussed the relationships between weak vector variational-like inequality problems and set-valued optimization problems. However, to the best of our knowledge, there are few papers discussing the solution relationships between set-valued optimization problems and strong vector variational-like inequality problems. Motivated by the works in [9] and [11], in this paper, we firstly introduce several kinds of generalized invexity for set-valued mappings and then prove that the solutions of the variational-like inequality problems are equivalent to the minima (ideal minima) of set-valued optimization problems.

The rest of the paper is organized as follows. In Section 2, we give some preliminaries and recall the main notions of contingent cone. In Section 3, the concept of contingent epiderivatives is introduced and, under determined conditions, its existence theory is also established. In Section 4, we present that the solutions of the variational-like inequality problems are equivalent to the minima (ideal minima) of set-valued optimization problems.

2 Preliminaries and notations

Throughout the paper, if not stated otherwise, let X be a real normed space, let A be a nonempty subset of X, and let Y be a real normed space partially ordered by a closed, convex and pointed cone DY. The points of origin of all real normed spaces are denoted by 0 X and 0 Y .

Let y 1 , y 2 Y, the orderings are defined in Y as follows:

y 1 y 2 y 2 y 1 D;

y 1 y 2 y 2 y 1 D.

Let AX, F:A 2 Y be a set-valued mapping. The graph, the epigraph and the domain of F are defined, respectively, by

gr F = { ( x , y ) X × Y : y F ( x ) } , epi F = { ( x , y ) X × Y : y F ( x ) + D } ,

and

domF= { x X : F ( x ) } .

We say that F is a D-convex set-valued mapping on A, if A is a convex set, and for all x,yA, all λ[0,1],

λF(x)+(1λ)F(y)F ( λ x + ( 1 λ ) y ) +D.

It is well known that if F is D-convex on A, then epiF is a convex subset in X×Y.

Proposition 2.1 (see [12])

Let F:X 2 Y . If epiF is a closed subset in X×Y, then F(x)+D is a closed set for each xX.

Definition 2.1 Let B be a non-empty subset of Y.

  1. (i)

    y 0 B is called a minimal point of B with respect to cone D if (B y 0 )D={ 0 Y };

  2. (ii)

    y 0 B is called an ideal minimal point of B with respect to cone D if B y 0 D.

The elements of all minimal points and ideal minimal points of B are denoted by MinB and IMinB, respectively.

Obviously, IMinBMinB if D is a pointed cone.

The following standard notions can be found in [6].

Definition 2.2 Let D be a closed, convex and pointed cone in Y.

  1. (i)

    The cone D is called Daniell cone if any decreasing sequence in Y having a lower bound converges to its infimum.

  2. (ii)

    A subset B of Y is said to be D-lower bounded (or be minorized) if there is an element yY such that By+D.

Proposition 2.2 Let D be a closed, convex and pointed cone in Y, non-empty subset BY. If MinB and B is the D-lower bounded for every minimal point, i.e.,

for each  y ¯ MinB,B y ¯ +D,
(2.1)

then MinB is a single-point set and MinB=IMinB.

Proof We prove this result by contradiction. Suppose that at least there are y 1 , y 2 MinB and y 1 y 2 . By condition (2.1), we have

y 1 y 2 +D,
(2.2)

and

y 2 y 1 +D.
(2.3)

From (2.2), (2.3) and the assumption y 1 y 2 , one has y 1 y 2 D{ 0 Y } and y 2 y 1 D{ 0 Y }, which contradicts D being a pointed cone, thus y 1 = y 2 = y ¯ . Namely, MinB is a singleton. Again from (2.1), we have

B y ¯ +D,

i.e.,

B y ¯ D,

this yields y ¯ IMinB. The proof is complete. □

Remark 2.1 We notice if MinB and (2.1) is fulfilled, then IMinB=MinB is a single-point set. Clearly, if IMinB, then BIMinB+D.

The following theorem is due to Borwein.

Theorem 2.1 (see [13])

Assume that D is a pointed, convex and Daniell cone, and let B be a closed subset of Y. If B is D-lower bounded (or is minorized), then MinB.

Definition 2.3 For AX, coneA and clA denote its cone hull and closure of A, respectively.

Furthermore,

coneA={λa:λ0,aA}.

Let AX, the contingent cone of A at x 0 is defined by

T(A, x 0 )= { u X : t n 0 + , u n u , n 0 N , n n 0 , x 0 + t n u n A } .

It is well known that if A is a convex set and x 0 A, then

T(A, x 0 )=clcone(A x 0 ).

We say that AX satisfies the property Ϝ, if for any xA and λ[0,1], one has λxA. Let 0 X A, we say that A satisfies the property Λ near the 0 X , if there exists neighborhood B( 0 X ,ε) such that for any xB( 0 X ,ε)(clAintA) and λ[0,1], one has λx(clAintA), where B( 0 X ,ε) denotes the ball centered at 0 X with radius ε.

Corollary 1 Let AX be a convex set, x 0 A, then A x 0 satisfies the property Ϝ.

Proof For any xA and t[0,1], t(x x 0 )+ x 0 =tx+(1t) x 0 , thereby, t(x x 0 )+ x 0 A follows immediately from A being a convex set, which implies t(x x 0 )A x 0 , the proof is complete. □

Proposition 2.3 Let AX, if A satisfies the properties Ϝ and Λ near the 0 X , then cone(clA)=cl(coneA).

Proof Obviously, 0 X cone(clA) and 0 X cl(coneA), so we consider u 0 X in the sequel. First, we prove the inclusion cone(clA)cl(coneA). Let ucone(clA), then there exist aclA, t>0 such that u=ta. Furthermore, following aclA, there exists { a n }A, a n a. For each n, set u n =t a n , then u n coneA and u n =t a n ta=u, i.e., u n u, this implies ucl(coneA).

For the contrary inclusion, we should only prove cone(clA) is a closed cone. Let { u n }cone(clA) and u n u. Next, we will prove ucone(clA). In fact, we can conclude that there exist x n X and x n 0 X such that u+ x n = u n cone(clA). So, for each n, there exist t n >0, a n clA such that

u+ x n = u n = t n a n .
(2.4)

On the one hand, if a n 0 X , then t n + as n+. For each n, we have

u t n + x n t n = a n clA.

For given ε>0, there exists n 0 such that a n B( 0 X ,ε), n n 0 . We divide it into two cases to discuss.

  1. (i)

    If { a n }clAintA as n n 0 , by assumption, we can conclude that there exists a subsequence { a n m }{ a n } such that

    a n m = λ n m a n ¯ 0 ,
    (2.5)

with λ n m (0,1] and n ¯ 0 n 0 . In fact, since t n a n u, so for any ε 1 >0, there exists n 1 n 0 such that t n a n u ε 1 as n n 1 . If for any given n ¯ 0 n 1 and λ(0,1],

λ a n ¯ 0 { a n },
(2.6)

then we can take some t n from { t n } such that t n ¯ 0 t n <1 as n n ¯ 0 . From (2.6) it follows that t n ¯ 0 t n a n ¯ 0 { a n } as n n ¯ 0 . Thus t n ( t n ¯ 0 t n a n ¯ 0 )u= t n ¯ 0 a n ¯ 0 u> ε 1 , a contradiction. So (2.5) holds. Again from t n a n u, so, t n m a n m = t n m λ n m a n ¯ 0 u. Furthermore, a n ¯ 0 is fixed, so t n m λ n m t 0 . That is, t n m λ n m = t 0 + t ¯ n m with t ¯ n m 0. Thus, t n m a n m =( t 0 + t ¯ n m ) a n ¯ 0 = t 0 a n ¯ 0 + t ¯ n m a n ¯ 0 t 0 a n ¯ 0 . From the uniqueness of limits, we have u= t 0 a n ¯ 0 with a n ¯ 0 clAintA, so ucone(clA).

  1. (ii)

    If { a n }intA, then for any n, ε n such that B( a n , ε n )A. From t n a n =u+ x n , one has a n u t n = x n t n and a n B( 0 X ,ε) as n is large enough. If there exists n 0 such that a n 0 u t n 0 = x n 0 t n 0 ε n 0 , then u t n 0 A, so ucone(clA). If, for any n, a n u t n = x n t n > ε n , then this implies there exist a ¯ n clAA and a ¯ n a n x n t n . However, t n a ¯ n u= t n a ¯ n t n a n + t n a n u t n x n t n + x n =2 x n 0. Thus, t n a ¯ n u. By a ¯ n clAA and (i), we conclude that ucone(clA).

On the other hand, if a n 0 X , then { t n } is bounded. Set t= sup n { t n }, then t t n and t R + . Dividing t by (2.4) and for each nN, set b n = t n a n t = u t + x n t . According to assumption A satisfying the property Ϝ, we get b n A and b n u t as n. Thus, u t clA, i.e., ucone(clA). From the above two parts, we get cone(clA) is a closed cone, so cl(coneA)cone(clA). The proof is complete. □

Corollary 2 Let A be a cone of X and satisfy the property Λ near the 0 X , then cone(clA)=clA.

Corollary 3 Let x 0 A, if A is a closed convex subset of X such that A x 0 satisfy the property Λ near the 0 X , then cone(A x 0 )=clcone(A x 0 ).

Proposition 2.4 (see Chapter 4 of [3])

Let x 0 AX, if A is star-shaped at x 0 , then T(A, x 0 )=cl(cone(A x 0 )).

Corollary 4 Let x 0 AX, if A is a closed convex subset such that A x 0 satisfy the property Λ near the 0 X , then T(A, x 0 )=cone(A x 0 ).

Proof Since x 0 A and A is a convex subset, we have T(A, x 0 )=cl(cone(A x 0 )). To apply Corollary 3, one has

cl ( cone ( A x 0 ) ) =cone(A x 0 ),

it is clear that T(A, x 0 )=cone(A x 0 ). □

3 The existence theory for contingent epiderivative

The aim of this section is to discuss the existence of the contingent epiderivative for a set-valued map defined from a real normed space to a real normed space. Let AX, F:A 2 Y be a set-valued mapping.

Definition 3.1 (see [1])

Let x 0 A and a pair ( x 0 , y 0 )grF be given. The contingent epiderivative DF( x 0 , y 0 ) of F at ( x 0 , y 0 ) is the single-valued mapping from X to Y defined by

epi ( D F ( x 0 , y 0 ) ) =T ( epi F , ( x 0 , y 0 ) ) .
(3.1)

Next, we establish DF( x 0 , y 0 ) in the following (3.6) and also prove it is a single-valued mapping.

Let a pair ( x 0 , y 0 )grF, for any xX, set

DF( x 0 , y 0 )(x)=IMin { y ( x , y ) T ( epi F , ( x 0 , y 0 ) ) } .
(3.2)

Notice that if AX is a closed convex set, F is D-convex and epiF is a closed subset in X×Y such that (epiF( x 0 , y 0 )) satisfies the property Λ near the ( 0 X , 0 Y ), then

T ( epi F , ( x 0 , y 0 ) ) = cone ( epi F { ( x 0 , y 0 ) } ) = cone ( ( gr F + { 0 } × D ) { ( x 0 , y 0 ) } ) = cone ( ( A × F ( A ) + { 0 } × D ) { ( x 0 , y 0 ) } ) = a A cone ( ( { a } × F ( a ) + { 0 } × D ) { ( x 0 , y 0 ) } ) ,

is a closed convex cone. So, for every (x,y)epi(DF( x 0 , y 0 )) and (x,y)( 0 X , 0 Y ), if (x,y)T(epiF,( x 0 , y 0 )), then there exist l>0, aA, vF(a) and dD such that

(x,y)=l(a x 0 ,v y 0 +d).
(3.3)

By (3.3), we have

x=l(a x 0 )cone(A x 0 ),
(3.4)

and

y=l(v y 0 +d)cone ( F ( x 0 + λ x ) y 0 + D ) ,
(3.5)

where λ= 1 l .

Otherwise, if (x,y)=( 0 X , 0 Y ), then we can conclude that a= x 0 , v= y 0 and d= y 0 . Therefore, formulas (3.4) and (3.5) are clearly established.

Remark 3.1 From the above discussion and Corollary 4, if AX is a closed convex set and A x 0 satisfies the property Λ near the 0 X , F:A 2 Y is D-convex and epiF is closed in X×Y such that (epiF( x 0 , y 0 )) satisfies the property Λ near the ( 0 X , 0 Y ), then DF( x 0 , y 0 ) can be defined on T(A, x 0 ).

Furthermore, we can conclude that formula (3.2) is equivalent to the following:

DF( x 0 , y 0 )(x)=IMin ( cone ( F ( x 0 + λ x ) y 0 + D ) ) ,
(3.6)

where xT(A, x 0 ) and λ0.

From Theorem 2.1, we can obtain the following existence theorem.

Theorem 3.1 Let AX be a closed convex set, let F:A 2 Y be D-convex, and let epiF be a closed subset in X×Y. Let ( x 0 , y 0 )grF. If for any xT(A, x 0 ) and λ0, the following conditions hold: (1) (F( x 0 +λx) y 0 +D) satisfies the properties Ϝ and Λ near the 0 Y ; (2) cone(F( x 0 +λx) y 0 +D) is D-lower bounded; (3) the cone D is Daniell cone; (4) cone(F( x 0 +λx) y 0 +D) satisfies (2.1). Then IMin(cone(F( x 0 +λx) y 0 +D)) is a single-point set.

Proof Since epiF a closed convex subset in X×Y, by Proposition 2.1, one has F( x 0 +λx)+D a closed set for each xT(A, x 0 ), so is (F( x 0 +λx) y 0 +D). By assumption (1) and Proposition 2.3, we can conclude that cone(F( x 0 +λx) y 0 +D) is a closed set. Considering Theorem 2.1 and given conditions (2) and (3), we have Min(cone(F( x 0 +λx) y 0 +D)). Combining Remark 2.1 and assumption (4), we can conclude IMin(cone(F( x 0 +λx) y 0 +D)) is a single-point set. □

Remark 3.2 From (3.6), we know that DF( x 0 , y 0 ) exists if IMin(cone(F( x 0 +λx) y 0 +D)).

The following theorem shows that our definition of the contingent epiderivative for a set-valued mapping is well defined.

Theorem 3.2 If all the conditions of Theorem  3.1 are fulfilled, then epi(DF( x 0 , y 0 ))=T(epiF,( x 0 , y 0 )).

Proof Combining (3.6) and Theorem 3.1, we get DF( x 0 , y 0 )(x) exists for each xT(A, x 0 ). Since (x,DF( x 0 , y 0 )(x))cone(epiF( x 0 , y 0 )), thus

epi ( D F ( x 0 , y 0 ) ) cone ( epi F { ( x 0 , y 0 ) } ) + { 0 } × D = cone ( epi F { ( x 0 , y 0 ) } ) = T ( epi F , ( x 0 , y 0 ) ) .

Now, let us show the converse inclusion. From the definition of the DF( x 0 , y 0 ), for any (x,y)T(epiF,( x 0 , y 0 )), we have

( x , y ) ( ( x , D F ( x 0 , y 0 ) ) + { 0 } × D ) gr ( D F ( x 0 , y 0 ) ) + { 0 Y } × D = epi ( D F ( x 0 , y 0 ) ) .

Therefore,

T ( epi F , ( x 0 , y 0 ) ) epi ( D F ( x 0 , y 0 ) ) .

The proof is complete. □

Next we show under appropriate assumptions that the contingent epiderivative is a strictly positive homogeneous and subadditive map in the case of A, a convex set, and epiF, a closed convex subset.

Definition 3.2 Let X be a real linear space and let Y be a real linear space partially ordered by a closed convex pointed cone DY. A map f:X 2 Y is called

  1. (a)

    strictly positive homogeneous if

    f(αx)=αf(x)for all α>0 and all xX,
  2. (b)

    subadditive if

    f( x 1 )+f( x 2 )f( x 1 + x 2 )+Dfor all  x 1 , x 2 X.

If the properties under (a) with α0 and (b) hold, then f is called sublinear.

Theorem 3.3 Let A be a closed convex set in X, x 0 A such that A x 0 satisfies the property Λ near the 0 X . Let D be a closed convex pointed cone in Y and ( x 0 , y 0 )grF such that epiF is a closed convex set and (epiF( x 0 , y 0 )) satisfies the property Λ near the ( 0 X , 0 Y ). If for all xT(A, x 0 ), DF( x 0 , y 0 )(x) exists, then DF( x 0 , y 0 )(x) is sublinear.

Proof We take any α>0 and any xT(A, x 0 ). Then we obtain

D F ( x 0 , y 0 ) ( α x ) = IMin { y ( α x , y ) T ( epi F , ( x 0 , y 0 ) ) } = IMin { y y ( cone ( F ( x 0 + λ α x ) y 0 + D ) ) } = IMin { α v α v ( cone ( F ( x 0 + λ α x ) y 0 + D ) ) } = α IMin { v v ( cone ( F ( x 0 + λ x ) y 0 + D ) ) } .

Thus DF( x 0 , y 0 ) is strictly positive homogeneous.

Next, for x 1 , x 2 T(A, x 0 ), we have ( x 1 ,DF( x 0 , y 0 )( x 1 ))T(epiF,( x 0 , y 0 )) and ( x 2 ,DF( x 0 , y 0 )( x 2 ))T(epiF,( x 0 , y 0 )). Since F is D-convex and epiF is a closed subset, then T(epiF,( x 0 , y 0 )) is a closed convex cone. Thus,

( x 1 + x 2 , D F ( x 0 , y 0 ) ( x 1 ) + D F ( x 0 , y 0 ) ( x 2 ) ) T ( epi F , ( x 0 , y 0 ) ) ,

implies

DF( x 0 , y 0 )( x 1 )+DF( x 0 , y 0 )( x 2 )cone ( F ( x 0 + λ ( x 1 + x 2 ) ) y 0 + D ) .

By Remark 2.1, we have

D F ( x 0 , y 0 ) ( x 1 ) + D F ( x 0 , y 0 ) ( x 2 ) cone ( F ( x 0 + λ ( x 1 + x 2 ) ) y 0 + D ) IMin ( cone ( F ( x 0 + λ ( x 1 + x 2 ) ) y 0 + D ) ) + D = D F ( x 0 , y 0 ) ( x 1 + x 2 ) + D .

The proof is complete. □

4 Relationships between vector variational-like inequalities and optimization problems

In this section, we restrict ourselves to dealing with relationships between two kinds of vector variational-like inequality problems and set-valued optimization problems.

The set-valued vector optimization problem under our consideration is

Min { F ( x ) : x A } ,
(4.1)

where F:X 2 Y , F(A)= x A F(x). We denote set-valued optimization problem (4.1) as (SOP).

Definition 4.1 Consider the above problem (SOP), let x 0 A, y 0 F( x 0 ).

  1. (i)

    A pair ( x 0 , y 0 )grF is called a minimal solution of F on A if (F(A) y 0 )D={ 0 Y };

  2. (ii)

    A pair ( x 0 , y 0 )grF is called an ideal minimal solution of F on A if (F(A) y 0 )D.

The sets of all minimal solutions and ideal minimal solutions of (SOP) are denoted by Min(F,A) and IMin(F,A), respectively.

In the following, we always assume that contingent epiderivative of F exists.

Definition 4.2 (see [14])

Let the set A be convex, let the set-valued mapping F:A 2 Y be D-convex. Let ( x 0 , y 0 )grF and let the contingent epiderivative DF( x 0 , y 0 ) of F at ( x 0 , y 0 ) exist.

  1. (i)

    A linear map L:XY, with L(x)DF( x 0 , y 0 )(x), for all xT(A, x 0 ) is called a subgradient of F at ( x 0 , y 0 ).

  2. (ii)

    The set F( x 0 , y 0 )={L:XY:L(x)DF( x 0 , y 0 )(x),xT(A, x 0 )} of all subgradients L of F at ( x 0 , y 0 ) is called the subdifferential of F at ( x 0 , y 0 ).

Definition 4.3 A set AX is said to be an invex set if there exists a function η:X×XX such that x,yA, λ[0,1], y+λη(x,y)A.

Throughout the paper, we always assume that A is an invex subset of X, the function η defined on A, i.e., η:A×AX and the subdifferential of F exists at every (x,y)grF.

Definition 4.4 Let F:A 2 Y be a set-valued mapping, ( x 0 , y 0 )grF. F is called strong pseudo-invex at ( x 0 , y 0 ) with respect to η on A if xA, yF(x) and LF( x 0 , y 0 ), y y 0 DLη(x, x 0 )D.

F is said to be strong pseudo-invex with respect to η on A if for every pair (x,y)grF, F is strong pseudo-invex at (x,y) with respect to η.

Definition 4.5 Let F:A 2 Y be a set-valued mapping, ( x 0 , y 0 )grF. F is called pseudo-invex at ( x 0 , y 0 ) with respect to η on A if xA, yF(x) and LF( x 0 , y 0 ), y y 0 DLη(x, x 0 )D.

F is said to be pseudo-invex with respect to η on A if for every pair (x,y)grF, F is pseudo-invex at (x,y) with respect to η.

Definition 4.6 Let F:A 2 Y be a set-valued mapping. F is said to be D-preinvex with respect to η on A if x,yA, λ[0,1], λF(x)+(1λ)F(y)F(y+λη(x,y))+D.

From the above definitions, we have the following useful proposition.

Proposition 4.1 Let F:A 2 Y be D-preinvex with respect to η on A, ( x 0 , y 0 )grF. If DF( x 0 , y 0 ) exists on cone(A x 0 ), then for all LF( x 0 , y 0 ), all xA, all yF(x), y y 0 L,η(x, x 0 )+D.

Proof For all xA, since F is D-preinvex with respect to η on A and x 0 A, for any λ[0,1], we obtain

(1λ)F( x 0 )+λF(x)F ( x 0 + λ η ( x , x 0 ) ) +D,

which implies that

y y 0 ( F ( x 0 + λ η ( x , x 0 ) ) + D y 0 ) /λ.
(4.2)

Thus, we get

y y 0 DF( x 0 , y 0 ) ( η ( x , x 0 ) ) +D.
(4.3)

It is clear that

xA,η(x, x 0 )domDF( x 0 , y 0 ).
(4.4)

By the definition of subdifferential F( x 0 , y 0 ), for all LF( x 0 , y 0 ), we have

L(x)DF( x 0 , y 0 )(x),xdomDF( x 0 , y 0 ),

that is,

DF( x 0 , y 0 )(x)L(x)+D.
(4.5)

Combining (4.3), (4.4) and (4.5), we can conclude, y y 0 L,η(x, x 0 )+D. □

Remark 4.1 It is easy to see that F is pseudo-invex with respect to η on A if F is D-preinvex with respect to η on A.

A vector variational inequality has been shown to be a useful tool in vector optimization. Some authors have proved the equivalence between them, see [9, 10, 15]. The vector variational-like inequality problem is a generalized form of the vector variational inequality problem, which was introduced and studied by [8].

Now, let us offer the following definitions.

A vector variational-like inequality problem (VVLI) is to find ( x 0 , y 0 )grF such that L,η(x, x 0 )D{ 0 Y } for all xA{ x 0 } and all LF( x 0 , y 0 ). A strong vector variational-like inequality problem (SVVLI) is to find ( x 0 , y 0 )grF such that L,η(x, x 0 )D for all xA and all LF( x 0 , y 0 ).

In the following, using the tools of non-smooth analysis and the concept of non-smooth (strong) vector pseudo-invexity, we shall obtain the stronger results than those of [11].

Theorem 4.1 Let x 0 A, ( x 0 , y 0 )grF. If the pair ( x 0 , y 0 ) solves the (SVVIL) and F is strong pseudo-invex at ( x 0 , y 0 ) with respect to η on A, then ( x 0 , y 0 ) is an ideal minimizer of the (SOP).

Proof Since the pair ( x 0 , y 0 ) solves the (SVVIL), we have that

L , η ( x , x 0 ) D,xA,LF( x 0 , y 0 ).
(4.6)

If ( x 0 , y 0 ) is not an ideal minimizer of the (SOP), then there exist x ¯ A, ( x ¯ , y ¯ )grF such that

y ¯ y 0 D.
(4.7)

Since F is strong pseudo-invex at ( x 0 , y 0 ) with respect to η on A, we get x ¯ A, LF( x 0 , y 0 ) such that

y ¯ y 0 D L , η ( x ¯ , x 0 ) D.
(4.8)

By (4.7) and (4.8), we obtain L,η( x ¯ , x 0 )D, which contradicts (4.6). □

In order to see the converse of the above theorem, we must impose stronger conditions, as can be observed in the following theorem.

Theorem 4.2 Suppose that F:A 2 Y andF is D-preinvex with respect to η on A, ( x 0 , y 0 )grF. If F( x 0 , y 0 )(F)( x 0 , y 0 ) and the pair ( x 0 , y 0 ) is an ideal minimizer of (SOP), then ( x 0 , y 0 ) solves the (SVVLI).

Proof Let ( x 0 , y 0 ) be an ideal minimizer of the (SOP), then we have

y y 0 D,xA,yF(x).
(4.9)

Since −F is D-preinvex with respect to η and F( x 0 , y 0 )(F)( x 0 , y 0 ), we have that xA, yF(x) and LF( x 0 , y 0 ) such that

y+ y 0 L , η ( x , x 0 ) +D,

which implies

L , η ( x , x 0 ) y y 0 +D.
(4.10)

Combining (4.9) and (4.10), we can conclude

L , η ( x , x 0 ) D+DD

for all xA and all LF( x 0 , y 0 ). That is, ( x 0 , y 0 ) solves the (SVVLI). □

Theorem 4.3 Suppose that F:A 2 Y andF is D-preinvex with respect to η on A, ( x 0 , y 0 )grF. If F( x 0 , y 0 )(F)( x 0 , y 0 ) and the pair ( x 0 , y 0 ) is a minimizer of (SOP), then ( x 0 , y 0 ) solves the (VVLI).

Proof Let ( x 0 , y 0 ) be a minimizer of the (SOP), then we have

y y 0 ( D { 0 Y } ) ,xA{ x 0 },yF(x).
(4.11)

If ( x 0 , y 0 ) does not solve the (VVIL), then there exist x ¯ A{ x 0 } and L ¯ F( x 0 , y 0 ) such that

L ¯ , η ( x ¯ , x 0 ) ( D { 0 Y } ) .
(4.12)

Since −F is D-preinvex with respect to η and F( x 0 , y 0 )(F)( x 0 , y 0 ), we have that xA, yF(x) and L ¯ F( x 0 , y 0 ) such that

y+ y 0 L ¯ , η ( x , x 0 ) +D.
(4.13)

Combining (4.12) and (4.13), for all y ¯ F( x ¯ ), one has

y ¯ + y 0 D{ 0 Y }+DD{ 0 Y },

which contradicts (4.11). The proof is complete. □

The converse case of the above theorem (see Theorem 4.4 of [11]) only requires F to be pseudo-invex.

References

  1. Jahn J, Raüh R: Contingent epiderivatives and set-valued optimization. Math. Methods Oper. Res. 1997, 46: 193–211. 10.1007/BF01217690

    Article  MathSciNet  Google Scholar 

  2. Aubin JP, Frankowska H: Set-Valued Analysis. Birkhäuser, Basel; 1990.

    Google Scholar 

  3. Jahn J, Khan AA: The existence of contingent epiderivatives for set-valued maps. Appl. Math. Lett. 2003, 16: 1179–1185. 10.1016/S0893-9659(03)90114-5

    Article  MathSciNet  Google Scholar 

  4. Rodríguez-Marín L, Sama M: About contingent epiderivatives. J. Math. Anal. Appl. 2007, 327: 745–762. 10.1016/j.jmaa.2006.04.060

    Article  MathSciNet  Google Scholar 

  5. Lalitha CS, Arora R: Weak Clarke epiderivative in set-valued optimization. J. Math. Anal. Appl. 2008, 342: 704–714. 10.1016/j.jmaa.2007.11.057

    Article  MathSciNet  Google Scholar 

  6. Chen GY, Jahn J: Optimality conditions for set-valued optimization problems. Math. Methods Oper. Res. 1998, 48: 187–200. 10.1007/s001860050021

    Article  MathSciNet  Google Scholar 

  7. Rodríguez-Marí L, Sama M: Variational characterization of the contingent epiderivative. J. Math. Anal. Appl. 2007, 335: 1374–1382. 10.1016/j.jmaa.2007.01.110

    Article  MathSciNet  Google Scholar 

  8. Ruiz-Garzón G, Osuna-Gómez R, Rufián-Lizana A: Relationships between vector variational-like inequality and optimization problems. Eur. J. Oper. Res. 2004, 157: 113–119. 10.1016/S0377-2217(03)00210-8

    Article  Google Scholar 

  9. Mishra SK, Wang SY: Vector variational-like inequalities and non-smooth vector optimization problems. Nonlinear Anal. 2006, 64: 1939–1945. 10.1016/j.na.2005.07.030

    Article  MathSciNet  Google Scholar 

  10. Rezaie M, Zafarani J: Vector optimization and variational-like inequalities. J. Glob. Optim. 2009, 43: 47–66. 10.1007/s10898-008-9290-1

    Article  MathSciNet  Google Scholar 

  11. Zeng J, Li SJ: On vector variational-like inequalities and set-valued optimization problems. Optim. Lett. 2011, 5: 55–69. 10.1007/s11590-010-0190-1

    Article  MathSciNet  Google Scholar 

  12. Hamel AH: A duality theory for set-valued functions I: Fenchel conjugation theory. Set-Valued Anal. 2009, 17: 153–182. 10.1007/s11228-009-0109-0

    Article  MathSciNet  Google Scholar 

  13. Borwein JM: On the existence of Pareto efficient points. Math. Oper. Res. 1983, 8: 64–73. 10.1287/moor.8.1.64

    Article  MathSciNet  Google Scholar 

  14. Baier J, Jahn J: Technical note on subdifferentials of set-valued maps. J. Optim. Theory Appl. 1999, 100: 233–240. 10.1023/A:1021733402240

    Article  MathSciNet  Google Scholar 

  15. Lee GM, Kim DS, Lee BS, Cho SJ: On vector variational inequality. Bull. Korean Math. Soc. 1996, 33(4):553–564.

    MathSciNet  Google Scholar 

Download references

Acknowledgements

This work was supported by the Fundamental Research Funds for the Central Universities, No. K5051370004.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Yanfei Chai.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

YC carried out the main preliminaries, the existence theory for contingent epiderivative and the relationships between vector variational-like inequalities and optimization problems. SL carried out the introduction and some preliminaries. YL participated in the design of the study and some preliminaries.

Authors’ original submitted files for images

Below are the links to the authors’ original submitted files for images.

Authors’ original file for figure 1

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 2.0 International License (https://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and permissions

About this article

Cite this article

Chai, Y., Liu, S. & Li, Y. The existence of contingent epiderivative for a set-valued mapping and vector variational-like inequalities. J Inequal Appl 2013, 352 (2013). https://doi.org/10.1186/1029-242X-2013-352

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/1029-242X-2013-352

Keywords