Algorithms for Common Solutions to Generalized Mixed Equilibrium Problems and Fixed Point Problems under Nonlinear Transformations in Banach Spaces ()
1. Introduction
Let H be a real Hilbert space, C be a nonempty closed convex subset of H, T be a mapping on C and
. Let
be a nonlinear mapping,
be a function and F be a bifunction from
to
, where
is the set of real numbers. Then, we consider the following generalized mixed equilibrium problem (for short, GMEP): finding
such that
(1)
The set of solutions of the GMEP is denoted by
(see [1] and the references therein). Here some special cases of the GMEP are stated as followings:
1) If
, then the GMEP becomes the following mixed equilibrium problem (for short, MEP):
(2)
which was studied by Ceng and Yao [2]. The set of solutions of the MEP is denoted by
.
2) If
and
, then the GMEP becomes the following equilibrium problem (for short, EP):
(3)
This general form of the EP was first considered by Nikaido and Isoda [3]. The MEP and EP play an important role in many fields, such as economics, physics, mechanics and engineering sciences. Also, the MEP and EP include many mathematical problems as particular cases, for example, mathematical programming problems, complementary problems, variational inequality problems, Nash equilibrium problems in noncooperative games, minimax inequality problems and fixed point problems. Because of their wide applicability, equilibrium problems and mixed equilibrium problems have been generalized in various directions for the past several years; see, for example, [2] [4] - [9].
3) If
and
, then the GMEP reduces to the following classical variational inequality problem (for short, VIP) [10] :
(4)
Since the VIP inception by Stampacchia [10] in 1964, it has received much attention due to its applications in a large variety of problems arising in structural analysis, economics, optimization, operations research and engineering sciences. Using the projection technique, one can easily show that is equivalent to the fixed-point problem; see, [7] [8] [11] [12] and the references therein.
Motivated by Ceng and Yao [2], Nikaido and Isoda [3] and Stampacchia [10], Peng and Yao [1] introduced the GMEP, which can be viewed as development and extension of the MEP, the EP and the VIP. It shows that the GMEP has applications in physics, economics, finance, transportation, network and structural analysis, therapy, image reconstruction, and elasticity. The GMEP includes special cases, MEPs, EPs, VIPs, fixed point problems, complementarity problems, optimization problems, Nash equilibrium problems in noncooperative games, etc (see e.g., [6] [7] [8] and the references contained in them). In other words, the GMEP is a unifying model for several problems arising in several areas of study. In general, the GMEP involves nonlinear equations and there are no known methods to obtain closed form solutions for them. Consequently, several methods are being deployed to approximate their solutions, assuming existence. A number of iterative methods have been utilized to solve equilibrium problems, generalized equilibrium problems and mixed equilibrium problems (see e.g., [2] [4] [5] [13] and the references therein).
Related to the GMEP, the problem of finding the fixed points for nonlinear mappings is the subject of current interest in functional analysis. It turns out that the fixed point theory for nonlinear mappings can be applied to several nonlinear problems such as zero point problems for monotone operators, convex feasibility problems, convex minimization problems, variational inequality and equilibrium problems, and so on; see [14] - [19] for more details.
At the same time, to construct a mathematical model which is as close as possible to a real complex problem, we often have to use more than one constraint. Solving such problems, we have to obtain some solution which is simultaneously the solution of two or more subproblems or the solution of one subproblem on the solution set of another subproblem. These subproblems can be given, for example, by two or more different variational inequality problems or two or more different fixed point problems. It is natural to construct a unified approach for these problems. In this direction, several authors have introduced some iterative schemes for finding a common solution of fixed-point problems for nonlinear mappings, equilibrium problems and variational problems; see, for example, [1] [2] [9] [12] [19] [20] and the references therein.
Recently, Takahashi [21] introduced a broad class of nonlinear mappings in a Banach space called k-demimetric mapping. This class mapping contains the classes of generalized hybrid mappings, k-strict pseudo-contractions, firmly-quasi-nonexpansive mappings, quasi-nonexpansive mappings and demicontractive mappings.
Definition 1.1 Let E be a smooth Banach space and let C be a nonempty, closed and convex subset of E. Let k be a real number with
. A mapping
with
is called k-demimetric if, for any
and
,
(5)
We give an example of a k-demimetric mapping which is not pseudo-contractive, hence it is not strictly pseudo-contractive.
Example 1.2 ( [22] ) Let H be the real line and
. Define T on C by
if
and
. Clearly, 0 is the only fixed point of T. Also, for
,
for any
. Thus T is demimetric.
In order to find a common solution of fixed point problems for an finite family of demimetric mappings and the variational inequality problems for a infinite family of inverse strongly monotone mappings in a Hilbert space, Takahashi [12] recently introduced and studied the following iterative algorithm:
![]()
where
is a finite family of kj-demimetric and demiclosed mappings, and
is a finite family of
-inverse strongly monotone mappings. Then he obtained a strong convergence theorem under some mild restrictions on the parameters.
Very recently, Akashi and Takahashi [14] proposed the following Mann’s type iteration for finding a common solution of fixed-point problems for an infinite family of demimetric mappings without assuming that demimetric mappings are commutative:
![]()
where
is an infinite family of kj-demimetric and demiclosed mappings. Then they obtained a weak convergence theorem under certain appropriate assumptions on the parameters.
Most very recently, Takahashi [15] also introduced the following iteration process for finding a common solution of fixed-point problems with an infinite family of demimetric mappings and the variational inequality problems with an infinite family of inverse strongly monotone mappings in a Hilbert space:
![]()
where
is an infinite family of kj-demimetric and demiclosed mappings,
is an infinite family of
-inverse strongly monotone mappings. Then they obtained a strong convergence theorem under some mild restrictions on the parameters.
On other hand, in order to find a common solution of equilibrium problems and the set of fixed point problems with generalized hybrid mappings, Alizadeh and Moradlou [23] introduced the following Ishikawa-like iteration process by applying the hybrid projection method:
![]()
where S is a generalized hybrid mapping and f is a bifunction satisfying (A1)-(A4). Then they obtained a strong convergence theorem under certain appropriate assumptions on the parameters.
Motivated and inspired by Takahashi [12], Akashi and Takahashi [14], Takahashi [15], Alizadeh and Moradlou [23], we put forward two questions:
1) Can these corresponding results in [12] [14] [15] [23] in Hilbert spaces be extended to the framework of Banach spaces (for example,
for
)?
2) Can we extend corresponding results in [12] [14] [15] [23] from finding a solution of the fixed point problems of generalized hybrid mappings or a common solution of the equilibrium problems and fixed point problems of generalized hybrid mappings to the more general and challenging problem for finding a common solution of the generalized mixed equilibrium problems and the fixed point problems of demimetric mappings under nonlinear transformations?
The purpose of this paper is to give the affirmative answers to these questions mentioned above. In this paper, we present a new iterative scheme for finding a common solution of the generalized mixed equilibrium problems and fixed point problems of demimetric mappings under nonlinear transformations in Banach spaces. Applications are also included. Our results improve essentially the corresponding results in [12] [14] [15] [23]. Further, some other results are also improved; see [9] [11] [16] [17] [18] [20] [24] [25].
2. Preliminaries
We denote E the real Banach space,
the dual of E, I the identity mapping on E, and
the set of positive integers. The expressions
and
denote the strong and weak convergence of the sequence
, respectively. The (normalized) duality mapping J from E to
is defined by
![]()
for all
, where
denotes the duality product. If E is a Hilbert space, then
, where I is the identity mapping on H.
The norm of a Banach space E is said to be Gâteaux differentiable if the limit
![]()
exists for all
on the unit sphere
. In this case, we say that E is smooth.
A Banach space E is said to be strictly convex if
whenever
and
. It is known that if E is strictly convex, then the duality mapping J is injective, that is,
and
imply
. It is known that E is reflexive if and only if J is surjective. Therefore, if E is a smooth, strictly convex and reflexive Banach space, then J is a single-valued bijection, see [26] for more details.
Definition 2.1 A mapping
is said to be:
1) nonexpansive if
for all
;
2) contractive if there exists a constant
such that
![]()
3)
-demicontractive if there exists a constant
such that
![]()
We use
to denote the collection of mappings T verifying the above inequality. That is
![]()
Let D be a nonempty subset of C. A sequence
of mappings of C into H is said to be stable on D (see [27] ) if
is a singleton for every
. It is clear that if
is stable on D, then
for all
and
.
Lemma 2.2 In a Hilbert space H, it holds for all
and
that
![]()
which can be extended to the more general situation: for all
,
, and
, we have
![]()
Lemma 2.3 ( [19] ) Let
be a sequence of real numbers such that there exists a subsequence
of
such that
for all
. Then there exists a nondecreasing sequence
such that
and the following properties are satisfied for all (sufficiently large) numbers
:
![]()
In fact,
.
Lemma 2.4 ( [28] ) Let
be a sequence of nonnegative numbers satisfying the property:
![]()
where
satisfy the restrictions:
1)
,
2)
,
3)
.
Then,
.
Lemma 2.5 ( [21] ) Let E be a smooth, strictly convex and reflexive Banach space and let
be a real number with
. Let U be an
-demimetric mapping of E into itself. Then
is closed and convex.
Lemma 2.6 ( [16] ) Let
be a metric projection from H on a nonempty closed convex subset C of H. Given
and
, then
if and only if there holds the relation
![]()
Recall that a mapping
is said to be
-inverse-strongly monotone (ism) if there exists a constant
such that
![]()
Lemma 2.7 ( [29] ) If
is
-ism and
is any constant in
, then the mapping
is nonexpansive.
For solving the generalized mixed equilibrium problem, let us assume that the bifunction
and the nonlinear mapping
satisfy the following conditions:
(A1)
for all
;
(A2) F is monotone, i.e.,
for all
;
(A3) for each fixed
,
is weakly upper semicontinuous;
(A4) for each fixed
,
is convex and lower semicontinuous;
(A5) for each
and
, there exists a bounded subset
and
such that, for any
,
![]()
(A6) C is a bounded set.
Lemma 2.8 [2] Let C be a nonempty, closed and convex subset of H and let
be a bifunction satisfying (A1)-(A4). Let
be a proper lower semicontinuous and convex function. Assume that either (B1) or (B2) holds. For
and
, define a mapping
as follows:
![]()
Then, the following conclusions hold:
1) For each
,
and
is single-valued;
2)
is a firmly nonexpansive mapping, i.e., for all
,
![]()
3)
;
4)
is closed and convex;
5)
for all
and
.
3. Main Results
Throughout the rest of this paper, we always assume the following:
1) H is a real Hilbert space, and C is a nonempty closed subspace of H;
2) E is a smooth, strictly convex and reflexive Banach space, and J is the duality mapping on E;
3) F is a bifunction from
to
satisfying (A1)-(A4);
4)
is a mapping defined as in Lemma 2.8;
5)
is an infinite family of
-ism mappings with
;
6)
is a lower semicontinuous and convex function with restrictions (B1) or (B2);
7)
is a bounded linear operator such that
and
is the adjoint operator of B;
8)
is an infinite family of
-demimetric and demiclosed mappings with
;
9)
;
10)
is stable on
.
Theorem 3.1 For any
, define a sequence
as follows:
(6)
where
,
and
satisfy the following conditions:
(i)
and
,
(ii)
and
,
(iii)
,
(iv)
and
,
(v)
.
Then the sequence
generated by (6) converges strongly to a point
, where
.
Proof. Set
for all
and
. Then we can prove that T is well defined. In fact, we have, for any
and
,
(7)
which implies
(8)
Thus,
![]()
Then we see the mapping
converges absolutely for each x in C.
Furthermore, define
for all
and
. Then we can prove that
is nonexpansive. Indeed, it follows that
is nonexpansive from (v), Lemma 2.7 and Lemma 2.8(2). We obtain from Lemma 2.5 that, for any
,
![]()
Thus
converges absolutely for each
.
Since
is nonexpansive, we have that
is closed and convex. Furthermore, we know from Lemma 2.5 that
is closed and convex for each
. Therefore, we have that
is nonempty, closed and convex (note that B is linear and continuous). Thus we have that
is well defined.
We derive from Lemma 2.8 that
(9)
Noting (8), we have
(10)
It follows from (9), (10) and (v) that
![]()
By induction, we obtain
![]()
which gives that the sequence
is bounded, so are
,
and
.
We obtain from (7) that
(11)
It follows from (9), (11) and Lemma 2.2 that
![]()
which means that
(12)
Case 1. Assume there exists some integer
such that
is decreasing for all
. In this case, we deduce that
exists. From (12), conditions (i), (ii), (iii) and (v), we deduce
(13)
and
(14)
From (6) and (13), we get that
![]()
Hence, we have
(15)
Since
is bounded, there exists a subsequence
of
satisfying
. Without loss of generality, we may also assume
(16)
Because B is bounded and linear, we see that
. This together with (13) implies
for each
. And hence,
.
Next let us prove that
. Noticing that a nonexpansive mapping
with
is 0-demimetric, then we have
![]()
This together with (14) and (15) implies, for any
, that
(17)
Consider a subsequence
of
corresponding to the sequence
. Since the subsequence
of
is bounded, we have that there exists a subsequence
of
such that
. For such r, we have from Lemma 2.8 (5) that
![]()
On the other hand, since
and
are Lipscitz, noting (17), we infer for any
that
(18)
Therefore, we obtain
.
It follows from (16) and Lemma 2.6 that
(19)
Putting
for all
, we have from (6) that
. Since
is stable on
, we then get by (9), (10) and Lemma 2.6 that
![]()
![]()
This together with Lemma 2.4 and (19) implies
as
.
Case 2: Suppose that there exists
of
such that
for all
. Then by Lemma 2.3, there exists a nondecreasing sequence
in
such that
(20)
Without loss of generality, there exists a subsequence
of
such that
for some
and
![]()
We show that
(21)
where
. To see this, we can first obtain
by a similar argument as in Case 1. Therefore, we deduce that
(22)
Like in Case 1, we can also get that
(23)
Observing that
![]()
we then find from (23) and (i) that
(24)
Putting
for all
, we obtain by Lemma 2.6, (9), (10) and (20) that
![]()
![]()
![]()
which means that
![]()
Noticing (22) and (24), we deduce
![]()
We can also obtain by (20) that
![]()
Consequently, we get
as
.
Remark 3.2 Theorem 3.1 extends, improves and develops Theorem 3.1 of Takahashi [12], Theorem 3.1 of Akashi and Takahashi [14], Theorem 3.1 of Takahashi [15] and Theorem 3.1 of Alizadeh and Moradlou [23] in the following aspects:
· Theorem 3.1 improves and develops corresponding results in [23] from generalized hybrid mappings to demimetric mappings;
· Theorem 3.1 extends, improves and develops corresponding results in [12] [14] and [15] from finding a common solution of fixed-point problems and the variational inequality problems in Hilbert spaces to the more general and challenging problem for finding a common solution of the generalized mixed equilibrium problems and the null point problems in Banach spaces;
· The proof of our Theorem 3.1 is very different from the proof of the ones given in [12] [14] [15] [23] ;
· The algorithm 6 is more advantageous and more flexible than the ones given in [12] [14] [15] [23]. Therefore, the new algorithm is expected to be widely applicable.
4. An Extension of Our Main Results
From Theorem 3.1, we deduce immediately the following results
Corollary 4.1 Suppose
. For
, define a sequence
as follows:
(25)
where
,
and
satisfy the following conditions:
1)
and
,
2)
and
,
3)
and
.
Then the sequence
generated by (25) converges strongly to a point
, where
.
Corollary 4.2 Let
be an infinite family of directed and demiclosed mappings. For
, define a sequence
as follows:
(26)
where
,
and
satisfy the following conditions:
1)
and
,
2)
and
,
3)
,
4)
and
,
5)
.
Then the sequence
generated by (26) converges strongly to a point
, where
.
Proof. Noticing that a directed mapping T with
is 0-demimetric, then we have the desired result due to Theorem 3.1.
5. Numerical Examples
In this section, we discuss the direct application of Theorem 3.1 on a typical example on a real line.
Example 5.1 Let
with the inner product defined by
for all
and the standard norm
. Let
,
,
and
be defined by
![]()
It is easy to check that
,
is an infinite family of
-demimetric and demiclosed mappings,
is an infinite family of
-ism mappings, and
is
-contractive on H and stable on
. Let
for all
, then we see
satisfies (A5).
Letting
for all
, we then see B is a bounded linear operator with its adjoint
. Note that
. Define a bifunction
by
![]()
We then find that F satisfies (A1)-(A4). So, by Lemma 2.8, we have
is nonempty and single-valued for each
. Hence, for any
, there exists
such that
![]()
which is equivalent to
![]()
After solving the above inequality, we get
, i.e.
.
Let us choose
,
,
,
,
, and
(choosing other values of these variables arbitrarily which satisfy the conditions of Theorem 3.1, the same convergence result also can be obtained). Then
,
,
,
,
,
and
satisfy all the conditions of Theorem 3.1. Then (6) can be rewrite as
![]()
It is not hard to estimate that
![]()
which shows
.
6. Conclusion
The present work has been aimed to theoretically establish a new iterative scheme for finding a common solution of the generalized mixed equilibrium problems with an infinite family of inverse strongly monotone mappings and the fixed point problems of demimetric mappings under nonlinear transformations in Banach spaces. Our results can be viewed as improvement, supplementation, development and extension of the corresponding results in some references to a great extent.
Acknowledgements
This research was supported by the Key Scientific Research Projects of Higher Education Institutions in Henan Province (20A110038).