Skip to main content

Weak and strong convergence of hybrid subgradient method for pseudomonotone equilibrium problem and multivalued nonexpansive mappings

Abstract

In this paper, we introduce a hybrid subgradient method for finding a common element of the set of solutions of a class of pseudomonotone equilibrium problems and the set of fixed points of a finite family of multivalued nonexpansive mappings in Hilbert space. The proposed method involves only one projection rather than two as in the existing extragradient method and the inexact subgradient method for an equilibrium problem. We establish some weak and strong convergence theorems of the sequences generated by our iterative method under some suitable conditions. Moreover, a numerical example is given to illustrate our algorithm and our results.

MSC:47H05, 47H09, 47H10.

1 Introduction

Let H be a real Hilbert space with inner product , and norm , respectively. Let K be a nonempty closed convex subset of H. Let F:K×KR be a bifunction, where denotes the set of real numbers. We consider the following equilibrium problem: Find xK such that

F(x,y)0,yK.
(1.1)

The set of solution of equilibrium problem is denoted by EP(F,K). It is well known that some important problems such as convex programs, variational inequalities, fixed point problems, minimax problems, and Nash equilibrium problem in noncooperative games and others can be reduced to finding a solution of the equilibrium problem (1.1); see [13] and the references therein.

Recall that a mapping T:KK is said to be nonexpansive if

TxTyxy,x,yK.

A subset KH is called proximal if for each xH, there exists an element yK such that

dist(x,K):=xy=inf { x z : z K } .

We denote by B(K), C(K), and P(K) the collection of all nonempty closed bounded subsets, nonempty compact subsets and nonempty proximal bounded subsets of K, respectively. The Hausdorff metric H on B(H) is defined by

H( K 1 , K 2 ):=max { sup x K 1 dist ( x , K 2 ) , sup y K 2 dist ( y , K 1 ) } , K 1 , K 2 B(H).

Let T:H 2 H be a multivalued mapping, of which the set of fixed points is denoted by Fix(T), i.e., Fix(T):={xTx:xK}. A multivalued mapping T:KB(K) is said to be nonexpansive if

H(Tx,Ty)xy,x,yK.
(1.2)

T is said to be quasi-nonexpansive if, for all pFix(T),

H(Tx,p)xp,xK.
(1.3)

Recently, the problem of finding a common element of the set of solutions of equilibrium problems and the set of fixed points of nonlinear mappings has become an attractive subject, and various methods have been extensively investigated by many authors. It is worth mentioning that almost all the existing algorithms for this problem are based on the proximal point method applied to the equilibrium problem combining with a Mann iteration to fixed point problems of nonexpansive mappings, of which the convergence analysis has been considered if the bifunction F is monotone. This is because the proximal point method is not valid when the underlying operator F is pseudomonotone. Another basic idea for solving equilibrium problems is the projection method. However, Facchinei and Pang [4] show that the projection method is not convergent for monotone inequality, which is a special case of monotone equilibrium problems. In order to obtain convergence of the projection method for equilibrium problems, Tran et al. [5] introduced an extragradient method for pseudomonotone equilibrium problems, which is computationally expensive because of the two projections defined onto the constrained set. Efforts for deducing the computational costs in computing the projection have been made by using penalty function methods or relaxing the constrained convex set by polyhedral convex ones; see, e.g., [615].

In 2011, Santos and Scheimberg [15] further proposed an inexact subgradient algorithm for solving a wide class of equilibrium problems that requires only one projection rather than two as in the extragradient method, and of which computational results show the efficiency of this algorithm in finite dimensional Euclidean spaces. On the other hand, iterative schemes for multivalued nonexpansive mappings are far less developed than those for nonexpansive mappings though they have more powerful applications in solving optimization problems; see, e.g., [1623] and the references therein.

In 2013, Eslamian [21] considered a proximal point method for nonspreading mappings and multivalued nonexpansive mappings and equilibrium problems. To be more precise, they proposed the following iterative method:

{ F ( u n , z ) + 1 r n y u n , u n x n 0 , y K , x n + 1 = α n u n + β n f n u n + γ n z n , n 1 ,
(1.4)

where T n = T n ( mod N ) , z n T n u n , α n + β n + γ n =1 for all n1 and f i , T i are finite families of nonspreading mappings and multivalued nonexpansive mappings for i=1,2,,N, respectively. Moreover, he further proved the weak and strong convergence theorems of the iterative sequences under the condition of monotone defined on a bifunction F.

In this paper, inspired and motivated by research going on in this area, we introduce a hybrid subgradient method for the pseudomonotone equilibrium problem and a finite family of multivalued nonexpansive mappings, which is defined in the following way:

{ w n ϵ n F ( x n , ) x n , u n = P K ( x n γ n w n ) , γ n = β n max { σ n , w n } , x n + 1 = α n x n + ( 1 α n ) z n , n 0 ,
(1.5)

where T n = T n ( mod N ) , z n T n u n , and { α n }, { β n }, { ϵ n }, and { σ n } are nonnegative real sequences.

Our purpose is not only to modify the proximal point iterative schemes (1.4) for the equilibrium problem to a hybrid subgradient method for a class of pseudomonotone equilibrium problems and a finite family of multivalued nonexpansive mappings, but also to establish weak and strong convergence theorems involving only one projection rather than two as in the extragradient method [5] and the inexact subgradient method [15] for the equilibrium problem. Our theorems presented in this paper improve and extend the corresponding results of [5, 15, 18, 21].

2 Preliminaries

Let K be a nonempty closed convex subset of a real Hilbert space H with inner product , and norm , respectively. For every point xH, there exists a unique nearest point in K, denoted by P K (x), such that

x P K ( x ) xy,yK.

Then P K is called the metric projection of H onto K. It is well known that P K is nonexpansive and satisfies the following properties:

x P K ( x ) , P K ( x ) y 0,xH,yK,
(2.1)
x y 2 x P K ( x ) 2 + y P K ( x ) 2 ,xH,yK.
(2.2)

Recall also that a bifunction F:K×KR is said to be

  1. (i)

    r-strongly monotone if there exists a number r>0 such that

    F(x,y)+F(y,x)r x y 2 ,x,yK;
  2. (ii)

    monotone on K if

    F(x,y)+F(y,x)0,x,yK;
  3. (iii)

    pseudomonotone on K with respect to xK if

    F(x,y)0F(y,x)0,yK.
    (2.3)

It is clear that (i) (ii) (iii), for every xK. Moreover, F is said to be pseudomonotone on K with respect to AK, if it is pseudomonotone on K with respect to every xA. When AK, F is called pseudomonotone on K.

The following example, taken from [15], shows that a bifunction may not be pseudomonotone on K, but yet is pseudomonotone on K with respect to the solution set of the equilibrium problem defined by F and K:

F(x,y):=2y|x|(yx)+xy|yx|,x,yR,K:=[1,1].

Clearly, EP(F,K)={0}. Since F(y,0)=0 for every yK, this bifunction is pseudomonotone on K with respect to the solution x =0. However, F is not pseudomonotone on K. In fact, both F(0.5,0.5)=0.25>0 and F(0.5,0.5)=0.25>0.

To study the equilibrium problem (1.1), we may assume that Δ is an open convex set containing K and the bifunction F:Δ×ΔR satisfy the following assumptions:

(C1) F(x,x)=0 for each xK and F(x,) is convex and lower semicontinuous on K;

(C2) F(,y) is weakly upper semicontinuous for each yK on the open set Δ;

(C3) F is pseudomonotone on K with respect to EP(F,K) and satisfies the strict paramonotonicity property, i.e., F(y,x)=0 for xEP(F,K) and yK implies yEP(F,K);

(C4) if { x n }K is bounded and ϵ n 0 as n, then the sequence { w n } with w n ϵ n F( x n ,) x n is bounded, where ϵ F(x,)x stands for the ϵ-subdifferential of the convex function F(x,) at x.

Throughout this paper, weak and strong convergence of a sequence { x n } in H to x are denoted by x n x and x n x, respectively. In order to prove our main results, we need the following lemmas.

Lemma 2.1 [24]

Let H be a real Hilbert space. For all x,yH, we have the following identity:

x y 2 = x 2 y 2 2xy,y.

Lemma 2.2 [25]

Let H be a real Hilbert space and α,β,γ[0,1] with α+β+γ=1. For all x,y,zH, we have the following identity:

α x + β y + γ z 2 = α x 2 + β y 2 + γ z 2 α β x y 2 α γ x z 2 β γ y z 2 .

Lemma 2.3 [26]

Let { a n } and { b n } be two sequences of nonnegative real numbers such that

a n + 1 a n + b n ,n0,

where n = 0 b n <. Then the sequence { a n } is convergent.

Lemma 2.4 [18]

Let K be a nonempty closed convex subset of a real Hilbert space H. Let T:KC(K) be a multivalued nonexpansive mapping. If x n q and lim n dist( x n ,T x n )=0, then qTq.

3 Weak convergence

Theorem 3.1 Let K be a nonempty closed convex subset of a Hilbert space H and F:K×KR be a bifunction satisfying (C1)-(C4). Let { T i } i = 1 N :KC(K) be a finite family of multivalued nonexpansive mappings such that Ω= i = 1 N Fix( T i )EP(F,K)ϕ and T i (q)={q} for i=1,2,,N and qΩ. For a given point x 0 K, 0<c< σ n <σ, { α n }, { β n }, and { ϵ n } are nonnegative sequences satisfying the following conditions:

  1. (i)

    α n [a,b](0,1);

  2. (ii)

    n = 0 β n =, n = 0 β n 2 <, and n = 0 β n ϵ n <.

Then the sequence { x n } generated by (1.5) converges weakly to x ¯ Ω.

Proof First, we show the existence of lim n x n p for every pΩ. It follows from (1.5) and Lemmas 2.1 and 2.2 that

x n + 1 p 2 = α n ( x n p ) + ( 1 α n ) ( z n p ) 2 = α n x n p 2 + ( 1 α n ) z n p 2 α n ( 1 α n ) x n z n 2 = α n x n p 2 + ( 1 α n ) dist ( z n , T n p ) 2 α n ( 1 α n ) x n z n 2 α n x n p 2 + ( 1 α n ) H ( T n u n , T n p ) 2 α n ( 1 α n ) x n z n 2 α n x n p 2 + ( 1 α n ) u n p 2 α n ( 1 α n ) x n z n 2 = α n x n p 2 + ( 1 α n ) ( x n p 2 u n x n 2 + 2 x n u n , p u n ) α n ( 1 α n ) x n z n 2 x n p 2 + 2 ( 1 α n ) x n u n , p u n α n ( 1 α n ) x n z n 2 .
(3.1)

By u n = P K ( x n γ n w n ) and (2.1), we have

x n u n ,p u n γ n w n ,p u n .
(3.2)

Using u n = P K ( x n γ n w n ) and x n K again, we obtain (note that γ n = β n max { σ n , w n } )

x n u n 2 = x n u n , x n u n γ n w n , x n u n γ n w n x n u n β n x n u n ,
(3.3)

which implies that x n u n β n . Substituting (3.2) into (3.1) yields

x n + 1 p 2 x n p 2 + 2 ( 1 α n ) γ n w n , p u n α n ( 1 α n ) x n z n 2 = x n p 2 + 2 ( 1 α n ) γ n w n , p x n + 2 ( 1 α n ) γ n w n , x n u n α n ( 1 α n ) x n z n 2 x n p 2 + 2 ( 1 α n ) γ n w n , p x n + 2 ( 1 α n ) γ n w n x n u n α n ( 1 α n ) x n z n 2 x n p 2 + 2 ( 1 α n ) γ n w n , p x n + 2 ( 1 α n ) β n 2 α n ( 1 α n ) x n z n 2 .
(3.4)

Since w n ϵ n F( x n ,) x n and F(x,x)=0 for all xK, we have

w n , p x n F ( x n , p ) F ( x n , x n ) + ϵ n F ( x n , p ) + ϵ n .
(3.5)

On the other hand, since pEP(F,K), i.e., F(p,x)0 for all xK, by the pseudomonotonicity of F with respect to p, we have F(x,p)0 for all xK. Replacing x by x n K, we get F( x n ,p)0. Then from (3.4) and (3.5), it follows that

x n + 1 p 2 x n p 2 + 2 ( 1 α n ) γ n F ( x n , p ) + 2 ( 1 α n ) γ n ϵ n + 2 ( 1 α n ) β n 2 α n ( 1 α n ) x n z n 2 x n p 2 + 2 ( 1 α n ) γ n ϵ n + 2 ( 1 α n ) β n 2 α n ( 1 α n ) x n z n 2 x n p 2 + 2 ( 1 α n ) γ n ϵ n + 2 ( 1 α n ) β n 2 .
(3.6)

Applying Lemma 2.3 to (3.6), by condition (ii), we obtain the existence of lim n x n p=d.

Now, we claim that lim sup n F( x n ,p)=0 for every pΩ. Indeed, since F is pseudomonotone on K and F(p, x n )0, we have F( x n ,p)0. From (3.6), we have

2 ( 1 α n ) γ n [ F ( x n , p ) ] x n p 2 x n + 1 p 2 + 2 ( 1 α n ) γ n ϵ n + 2 ( 1 α n ) β n 2 .
(3.7)

Summing up (3.7) for every n, we obtain

0 2 n = 0 ( 1 α n ) γ n [ F ( x n , p ) ] x 0 p 2 + 2 n = 0 γ n ϵ n + 2 n = 0 β n 2 < + .
(3.8)

By the assumption (C4), we can find a real number w such that w n w for every n. Setting L:=max{σ,w}, where σ is a real number such that 0< σ n <σ for every n, it follows from (i) that

0 2 ( 1 b ) L n = 0 β n [ F ( x n , p ) ] 2 n = 0 ( 1 α n ) γ n [ F ( x n , p ) ] < + ,

which implies that

n = 0 β n [ F ( x n , p ) ] <+.
(3.9)

Combining with F( x n ,p)0 and n = 0 β n =, we can deduced that lim sup n F( x n ,p)=0 as desired.

Next, we show that any weak subsequential limit of the sequence of { x n } is an element of Ω= i = 1 N Fix( T i )EP(F,K). To do this, suppose that { x n i } is a subsequence of { x n }. For simplicity of notation, without loss of generality, we may assume that x n i x ¯ as i. By convexity, K is weakly closed and hence x ¯ K. Since F(,p) is weakly upper semicontinuous for pΩ, we have

F ( x ¯ , p ) lim sup i F ( x n i , p ) = lim i F ( x n i , p ) = lim sup n F ( x n , p ) = 0 .
(3.10)

By the pseudomonotonicity of F with respect to p and F(p, x ¯ )0, we obtain F( x ¯ ,p)0. Thus F( x ¯ ,p)=0. Moreover, by the assumption (C3), we can deduce that x ¯ is a solution of EP(F,K). On the other hand, it follows from (3.3) and condition (ii) that

lim n x n u n =0.
(3.11)

From (3.6) and conditions (i)-(ii), we have

α n (1 α n ) x n z n 2 x n p 2 x n + 1 p 2 +2(1 α n ) γ n ϵ n +2(1 α n ) β n 2 ,

taking the limit as n yields

lim n x n z n =0,
(3.12)

and thus

lim n dist( x n , T n u n ) lim n x n z n =0.
(3.13)

Using (1.5) again, we have

lim n x n + 1 x n = lim n (1 α n ) x n z n =0.
(3.14)

It follows that

lim n x n + i x n =0,i=1,2,,N.
(3.15)

Note that

u n + 1 u n u n + 1 x n + 1 + x n + 1 x n + x n u n .

Combining (3.11) and (3.14), we obtain

lim n u n + 1 u n =0.
(3.16)

This also implies that

lim n u n + i u n =0,i=1,2,,N.
(3.17)

Observe that

dist ( u n , T n + i u n ) u n x n + x n x n + i + dist ( x n + i , T n + i u n + i ) + H ( T n + i u n + i , T n + i u n ) u n x n + x n x n + i + dist ( x n + i , T n + i u n + i ) + u n + i u n .

Together with (3.11), (3.13), (3.15), and (3.17), we have

lim n dist( u n , T n + i u n )=0,i=1,2,,N,
(3.18)

which implies that the sequence

i = 1 N { dist ( u n , T n + i u n ) } n 0 0as n.
(3.19)

For i=1,2,,N, we note also that

{ dist ( u n , T i u n ) } n 0 = { dist ( u n , T n + ( i n ) u n ) } n 0 = { dist ( u n , T n + i n u n ) } n 0 i = 1 N { dist ( u n , T n + i u n ) } n 0 ,

where in= i n (modN) and i n {1,2,,N}. Therefore, we have

lim n dist( u n , T i u n )=0,i=1,2,,N.
(3.20)

Similarly, for i=1,2,,N, we obtain

dist ( x n , T i x n ) x n u n + dist ( u n , T i u n ) + H ( T i u n , T i x n ) 2 x n u n + dist ( u n , T i u n ) .

It follows from (3.11) and (3.20) that

lim n dist( x n , T i x n )=0,i=1,2,,N.
(3.21)

Applying Lemma 2.4 to (3.21), we can deduce that x ¯ Fix( T i ) for i=1,2,,N and hence x ¯ Ω.

Finally, we prove that { x n } converges weakly to an element of Ω. Indeed to verify that the claim is valid it is sufficient to show that ω w ( x n ) is a single point set, where ω w ( x n )={xH: x n i x} for some subsequence { x n i } of { x n }. Indeed since { x n } is bounded and H is reflexive, ω w ( x n ) is nonempty. Taking w 1 , w 2 ω w ( x n ) arbitrarily, let { x n k } and { x n j } be subsequences of { x n } such that x n k w 1 and x n j w 2 , respectively. Since lim n x n p exists for all pΩ and w 1 , w 2 Ω, we see that lim n x n w 1 and lim n x n w 2 exist. Now let w 1 w 2 , then by Opial’s property,

lim n x n w 1 = lim k x n k w 1 < lim k x n k w 2 = lim n x n w 2 = lim j x n j w 2 < lim j x n j w 1 = lim n x n w 1 ,

which is a contradiction. Therefore, w 1 = w 2 . This shows that ω w ( x n ) is a single point set, i.e., x n x ¯ . This completes the proof. □

Theorem 3.2 Let K be a nonempty closed convex subset of a Hilbert space H and F:K×KR be a bifunction satisfying (C1)-(C4). Let T:KC(K) be a multivalued nonexpansive mapping such that Ω=Fix(T)EP(F,K)ϕ and T(q)={q} for all qΩ. For a given point x 0 K, 0<c< σ n <σ, let { x n } be defined by

{ w n ϵ n F ( x n , ) x n , u n = P K ( x n γ n w n ) , γ n = β n max { σ n , w n } , x n + 1 = α n x n + ( 1 α n ) z n , n 0 ,

where z n T u n , { α n }, { β n }, and { ϵ n } are nonnegative sequences satisfying the following conditions:

  1. (i)

    α n [a,b](0,1);

  2. (ii)

    n = 0 β n =, n = 0 β n 2 <, and n = 0 β n ϵ n <.

Then the sequence { x n } converges weakly to x ¯ Ω.

Proof Putting N=1, then T i =T, a single multivalued nonexpansive mapping, and the conclusion follows immediately from Theorem 3.1. This completes the proof. □

4 Strong convergence

To obtain strong convergence results, we either add the control condition lim n α n = 1 2 , or we remove the condition T(q)={q} for all qΩ and adjust the nonempty compact subset C(K) to a proximal bounded subset P(K) of K as follows.

Theorem 4.1 Let K be a nonempty closed convex subset of a Hilbert space H and F:K×KR be a bifunction satisfying (C1)-(C4). Let { T i } i = 1 N :KC(K) be a finite family of multivalued nonexpansive mappings such that Ω= i = 1 N Fix( T i )EP(F,K)ϕ and T i (q)={q} for i=1,2,,N and qΩ. For a given point x 0 K, 0<c< σ n <σ, let { x n } be defined by

{ w n ϵ n F ( x n , ) x n , u n = P K ( x n γ n w n ) , γ n = β n max { σ n , w n } , x n + 1 = α n x n + ( 1 α n ) z n , n 0 ,
(4.1)

where T n = T n ( mod N ) , z n T n u n , { α n }, { β n }, and { ϵ n } are nonnegative sequences satisfying the following conditions:

  1. (i)

    α n [a,b](0,1) and lim n α n = 1 2 ;

  2. (ii)

    n = 0 β n =, n = 0 β n 2 <, and n = 0 β n ϵ n <.

Then the sequence { x n } generated by (4.1) converges strongly to x Ω.

Proof By a similar argument to the proof of Theorem 3.1 and (2.2), we have

z n P Ω ( x n ) 2 z n x n 2 x n P Ω ( x n ) 2 .
(4.2)

It follows from (4.2) that

x n + 1 P Ω ( x n + 1 ) 2 α n ( x n P Ω ( x n ) ) + ( 1 α n ) ( z n P Ω ( x n ) ) 2 α n x n P Ω ( x n ) 2 + ( 1 α n ) z n P Ω ( x n ) 2 ( 2 α n 1 ) x n P Ω ( x n ) 2 + ( 1 α n ) z n x n 2 .
(4.3)

Combining (3.12), lim n α n = 1 2 , and the boundedness of the sequence { x n P Ω ( x n )}, we obtain

lim n x n + 1 P Ω ( x n + 1 ) =0.
(4.4)

By the assumptions (C1) and (C2), the set Ω is convex (see the proof of Theorem 1 in [27]). For all m>n, we have 1 2 ( P Ω ( x m )+ P Ω ( x n ))Ω, and therefore

P Ω ( x m ) P Ω ( x n ) 2 = 2 x m P Ω ( x m ) 2 + 2 x m P Ω ( x n ) 2 4 x m 1 2 ( P Ω ( x m ) + P Ω ( x n ) ) 2 2 x m P Ω ( x m ) 2 + 2 x m P Ω ( x n ) 2 4 x m P Ω ( x m ) 2 = 2 x m P Ω ( x n ) 2 2 x m P Ω ( x m ) 2 .
(4.5)

Using (3.6) with p= P Ω ( x n ), we have

x m P Ω ( x n ) 2 x m 1 P Ω ( x n ) 2 + 2 ( 1 α m 1 ) γ m 1 ϵ m 1 + 2 ( 1 α m 1 ) β m 1 2 x m 2 P Ω ( x n ) 2 + η m 1 + η m 2 x n P Ω ( x n ) 2 + j = n m 1 η j ,
(4.6)

where η j =2(1 α j ) γ j ϵ j +2(1 α j ) β j 2 . It follows from (4.5) and (4.6) that

P Ω ( x m ) P Ω ( x n ) 2 2 x n P Ω ( x n ) 2 +2 j = n m 1 η j 2 x m P Ω ( x m ) 2 .
(4.7)

Together with (4.4) and j = n m 1 η j <, this implies that { P Ω ( x n )} is a Cauchy sequence. Hence { P Ω ( x n )} strongly converges to some point x Ω. Moreover, we obtain

x = lim i P Ω ( x n i )= P Ω ( x ¯ )= x ¯ ,
(4.8)

which implies that P Ω ( x n ) x = x ¯ Ω. Then, from (3.21), (4.4), and (4.8), we can conclude that x n x . This completes the proof. □

Theorem 4.2 Let K be a nonempty closed convex subset of a Hilbert space H and F:K×KR be a bifunction satisfying (C1)-(C4). Let { T i } i = 1 N :KP(K) be a finite family of multivalued mappings such that P T i is nonexpansive, where P T i :={y T i x:dist(x, T i x)=xy} and Ω= i = 1 N Fix( T i )EP(F,K)ϕ. For a given point x 0 K, 0<c< σ n <σ, let { x n } be defined by

{ w n ϵ n F ( x n , ) x n , u n = P K ( x n γ n w n ) , γ n = β n max { σ n , w n } , x n + 1 = α n x n + ( 1 α n ) z n , n 0 ,
(4.9)

where T n = T n ( mod N ) , z n P T n u n , { α n }, { β n }, and { ϵ n } are nonnegative sequences satisfying the following conditions:

  1. (i)

    α n [a,b](0,1);

  2. (ii)

    n = 0 β n =, n = 0 β n 2 <, and n = 0 β n ϵ n <.

Then the sequence { x n } converges strongly to x Ω.

Proof Taking pΩ, then P T n (p)={p}. By substituting P T instead of T and similar argument as (3.21) in the proof of Theorem 3.1 we obtain

lim n dist ( x n , T i ( x n ) ) lim n dist ( x n , P T i ( x n ) ) =0.
(4.10)

By compactness of K, there exists a subsequence { x n k } of { x n } such that lim k x n k = x , for some x K. Since P T i is nonexpansive for i=1,2,,N, we have

dist ( x , T i ( x ) ) dist ( x , P T i ( x ) ) x x n k + dist ( x n k , P T i ( x n k ) ) + H ( P T i ( x n k ) , P T i ( x ) ) 2 x x n k + dist ( x n k , P T i ( x n k ) ) .
(4.11)

It follows from (4.10) and (4.11) that

lim k dist ( x , T i ( x ) ) =0,
(4.12)

which implies that x i = 1 N Fix( T i ). Since { x n k } converges strongly to x and lim n x n x exists (as in the proof of Theorem 3.1), we find that { x n } converges strongly to x . This completes the proof. □

In addition, we supply an example and numerical results to illustrate our method and the main results of this paper.

Example 4.1 Let H=R and K:=[0,1] with usual metric. Consider the nonsmooth equilibrium problem defined by the bifunction

F(x,y)=2xy(yx)+xy|yx|,x,yK.

Clearly, F is pseudomonotone on K. Note that F(x,) is convex for xK and F(x,)x=[ x 2 ,3 x 2 ] by taking ϵ n =0 for all nN.

(i) Let Tx:=[ x 6 , x 2 ] defined on K:=[0,1]. Note that T is a multivalued nonexpansive mapping and Fix(T)EP(F,K)={0}. Setting N=1, σ n =1, α n = 1 2 , β n = 1 n , and x n x 10 4 as stop criteria, we obtain the results of algorithm (4.1) with different initial points in Table 1.

Table 1 Numerical results for an initial point x 0 =0.2,0.5,0.8

(ii) Let Tx:=[2x,x] defined on [0,) 2 R . Note that T is not nonexpansive but P T x={x} is nonexpansive for all x[0,). Indeed, for each uTx, u=ax, 1a2, choose v=ay. Then

|uv|= | a x ( a y ) | =a|xy|2|xy|=H(Tx,Ty).

On the other hand, for any x, we have 0[0,) and T0={0}. It follows that Fix(T)EP(F,K)={0}. Setting N=1, σ n =1, α n = 1 4 , β n = 2 n , and x n x 10 4 as stop criteria, we obtain the results of algorithm (4.9) with different initial points to be found in Table 2.

Table 2 Numerical results for an initial point x 0 =0.2,0.5,0.8

The computations are performed by Matlab R2008a running on a PC Desktop Intel(R) Core(TM) i3-2330M, CPU @2.20 GHz, 790 MHz, 1.83 GB, 2 GB RAM.

Remark 4.1 Our hybrid subgradient method improves the extragradient method of Tran et al. [5] and the inexact subgradient algorithm of Santos and Scheimberg [15] for an equilibrium problem in deducing the computational costs of an iterative process.

Remark 4.2 Our results generalize the results of Eslamian [21], a proximal point method for an equilibrium problem, to a hybrid subgradient method for a pseudomonotone equilibrium problem.

References

  1. Blum E, Oettli W: From optimization and variational inequality to equilibrium problems. Math. Stud. 1994, 63: 127–149.

    MathSciNet  Google Scholar 

  2. Auslender A, Teboulle M, Ben-Tiba S: A logarithmic quadratic proximal method for variational inequalities. Comput. Optim. Appl. 1999, 12: 31–40. 10.1023/A:1008607511915

    Article  MathSciNet  Google Scholar 

  3. Combettes PL, Hirstoaga SA: Equilibrium programming in Hilbert spaces. J. Nonlinear Convex Anal. 2005, 6: 117–136.

    MathSciNet  Google Scholar 

  4. Facchinei F, Pang JS: Finite-Dimensional Variational Inequalities and Complementarity Problems. Springer, New York; 2003.

    MATH  Google Scholar 

  5. Tran QD, Muu LD, Nguyen VH: Extragradient algorithms extended to equilibrium problems. Optimization 2008, 57: 749–776. 10.1080/02331930601122876

    Article  MathSciNet  Google Scholar 

  6. Muu LD, Oettli W: Convergence of an adaptive penalty scheme for finding constrained equilibria. Nonlinear Anal. TMA 1992, 18: 1159–1166. 10.1016/0362-546X(92)90159-C

    Article  MathSciNet  Google Scholar 

  7. Anh PN: A logarithmic quadratic regularization method for solving pseudo-monotone equilibrium problem. Acta Math. Vietnam. 2009, 34: 183–200.

    MathSciNet  Google Scholar 

  8. Nguyen TTV, Strodiot JJ, Nguyen VH: The interior proximal extragradient method for solving equilibrium problems. J. Glob. Optim. 2009, 40: 175–192.

    Article  MathSciNet  Google Scholar 

  9. Anh PN, Kim JK: Outer approximation algorithms for pseudomonotone equilibrium problems. Comput. Math. Appl. 2011, 61: 2588–2595. 10.1016/j.camwa.2011.02.052

    Article  MathSciNet  Google Scholar 

  10. Wen D-J: Strong convergence theorems for equilibrium problems and k -strict pseudocontractions in Hilbert spaces. Abstr. Appl. Anal. 2011., 2011: Article ID 276874 10.1155/2011/276874

    Google Scholar 

  11. Qin X, Cho SY, Kang SM: An extragradient-type method for generalized equilibrium problems involving strictly pseudocontractive mappings. J. Glob. Optim. 2011, 49: 679–693. 10.1007/s10898-010-9556-2

    Article  MathSciNet  Google Scholar 

  12. Anh PN, Kim JK, Nam JM: Strong convergence of an extragradient method for equilibrium problems and fixed point problems. J. Korean Math. Soc. 2012, 49: 187–200. 10.4134/JKMS.2012.49.1.187

    Article  MathSciNet  Google Scholar 

  13. Wen D-J, Chen Y-A: General iterative methods for generalized equilibrium problems and fixed point problems of k -strict pseudo-contractions. Fixed Point Theory Appl. 2012., 2012: Article ID 125

    Google Scholar 

  14. Vuong PT, Strodiot JJ, Nguyen VH: Extragradient methods and linesearch algorithms for solving Ky Fan inequalities and fixed point problems. J. Optim. Theory Appl. 2012, 155: 605–627. 10.1007/s10957-012-0085-7

    Article  MathSciNet  Google Scholar 

  15. Santos P, Scheimberg S: An inexact subgradient algorithm for equilibrium problems. Comput. Appl. Math. 2011, 30: 91–107.

    MathSciNet  Google Scholar 

  16. Ceng LC, Yao JC: Hybrid viscosity approximation schemes for equilibrium problems and fixed point problems of infinitely many nonexpansive mappings. Appl. Math. Comput. 2008, 198: 729–741. 10.1016/j.amc.2007.09.011

    Article  MathSciNet  Google Scholar 

  17. Song Y, Wang H: Convergence of iterative algorithms for multivalued mappings in Banach spaces. Nonlinear Anal. 2009, 70: 1547–1556. 10.1016/j.na.2008.02.034

    Article  MathSciNet  Google Scholar 

  18. Eslamian M, Abkar A: One-step iterative process for a finite family of multivalued mappings. Math. Comput. Model. 2011, 54: 105–111. 10.1016/j.mcm.2011.01.040

    Article  MathSciNet  Google Scholar 

  19. Shahzad N, Zegeye H: On Mann and Ishikawa iteration schemes for multivalued maps in Banach space. Nonlinear Anal. 2009, 71: 838–844. 10.1016/j.na.2008.10.112

    Article  MathSciNet  Google Scholar 

  20. Qin X, Cho YJ, Kang SM: Viscosity approximation methods for generalized equilibrium problems and fixed point problems with applications. Nonlinear Anal. 2010, 72: 99–112. 10.1016/j.na.2009.06.042

    Article  MathSciNet  Google Scholar 

  21. Eslamian M: Convergence theorems for nonspreading mappings and nonexpansive multivalued mappings and equilibrium problems. Optim. Lett. 2013, 7: 547–557. 10.1007/s11590-011-0438-4

    Article  MathSciNet  Google Scholar 

  22. Wen D-J: Projection methods for generalized system of nonconvex variational inequalities with different nonlinear operators. Nonlinear Anal. 2010, 73: 2292–2297. 10.1016/j.na.2010.06.010

    Article  MathSciNet  Google Scholar 

  23. Long XJ, Quan J, Wen D-J: Proper efficiency for set-valued optimization problems and vector variational-like inequalities. Bull. Korean Math. Soc. 2013, 50: 777–786. 10.4134/BKMS.2013.50.3.777

    Article  MathSciNet  Google Scholar 

  24. Marino G, Xu HK: Weak and strong convergence theorems for strict pseudo-contractions in Hilbert space. J. Math. Anal. Appl. 2007, 329: 336–346. 10.1016/j.jmaa.2006.06.055

    Article  MathSciNet  Google Scholar 

  25. Osilike MO, Igbokwe DI: Weak and strong convergence theorems for fixed points of pseudocontractions and solutions of monotone type operator equations. Comput. Math. Appl. 2000, 40: 559–567. 10.1016/S0898-1221(00)00179-6

    Article  MathSciNet  Google Scholar 

  26. Xu HK: Viscosity approximation methods for nonexpansive mappings. J. Math. Anal. Appl. 2004, 298: 279–291. 10.1016/j.jmaa.2004.04.059

    Article  MathSciNet  Google Scholar 

  27. Muu LD: Stability property of a class of variational inequalities. Math. Oper.forsch. Stat., Ser. Optim. 1984, 15: 347–351.

    Google Scholar 

Download references

Acknowledgements

The author is grateful to the anonymous referees for valuable remarks suggestions which helped him very much in improving this manuscript. This work was supported by the National Science Foundation of China (11471059, 11271388), Basic and Advanced Research Project of Chongqing (cstc2014jcyjA00037) and Science and Technology Research Project of Chongqing Municipal Education Commission (KJ1400618).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Dao-Jun Wen.

Additional information

Competing interests

The author declares to have no competing interests.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0), which permits use, duplication, adaptation, distribution, and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wen, DJ. Weak and strong convergence of hybrid subgradient method for pseudomonotone equilibrium problem and multivalued nonexpansive mappings. Fixed Point Theory Appl 2014, 232 (2014). https://doi.org/10.1186/1687-1812-2014-232

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/1687-1812-2014-232

Keywords