Skip to main content

On split common solution problems: new nonlinear feasible algorithms, strong convergence results and their applications

Abstract

In this paper, we study and give examples for classes of generalized contractive mappings. We establish some new strong convergence theorems of feasible iterative algorithms for the split common solution problem (SCSP) and give some applications of these new results.

MSC:47H06, 47J25, 47H09, 65K10.

1 Introduction and preliminaries

Let K be a closed convex subset of a real Hilbert space H with the inner product , and the norm . The following inequalities are known and useful.

  • x + y 2 y 2 +2x,x+y;

  • x y 2 = x 2 + y 2 2x,y for all x,yH;

  • α x + ( 1 α ) y 2 =α x 2 +(1α) y 2 α(1α) x y 2 for all x,yH and α[0,1].

For each point xH, there exists a unique nearest point in K, denoted by P K x, such that

x P K xxyfor all yK.

The mapping P K is called the metric projection from H onto K. It is well known that P K has the following properties:

  1. (i)

    xy, P K x P K y P K x P K y 2 for every x,yH.

  2. (ii)

    For xH and zK, z= P K xxz,zy0 for all yK.

  3. (iii)

    For xH and yK,

    y P K x 2 + x P K x 2 x y 2 .
    (1.1)

Let H 1 and H 2 be two Hilbert spaces. Let A: H 1 H 2 and A : H 2 H 1 be two bounded linear operators. A is called the adjoint operator (or adjoint) of A if

Az,w= z , A w for all z H 1  and w H 2 .

It is known that the adjoint operator of a bounded linear operator on a Hilbert space always exists and is bounded linear and unique. Moreover, it is not hard to show that if A is an adjoint operator of A, then A= A . The symbols and are used to denote the sets of positive integers and real numbers, respectively.

Let H 1 and H 2 be two real Hilbert spaces. Let C be a closed convex subset of H 1 and K be a closed convex subset of H 2 . Let T:CC with F(T) and S:KK with F(S) be two mappings. Let A: H 1 H 2 be a bounded linear operator. The mathematical model of the split common solution problem (SCSP in short) is defined as follows:

(SCSP)Find pC such that Tp=p and u:=ApK satisfying Su=u.

In fact, SCSP contains several important problems as special cases and many authors have studied and introduced some new iterative algorithms for SCSP and presented some strong and weak convergence theorems for SCSP; see, for instance, [124] and the references therein. Motivated and inspired by their works, in this paper, we study and establish new strong convergence results by using new iterative algorithms of SCSP for pseudocontractive mappings and k-demicontractive mappings in Hilbert spaces.

The paper is divided into four sections. In Section 2, we study and give examples for classes of generalized contractive mappings. Some new strong convergence theorems of feasible iterative algorithms for SCSP are established in Section 3. Finally, some applications and further remarks for our new results are given in Section 4. Consequently, in this paper, some of our results are original and completely different from these known related results in the literature.

2 Classes of generalized contractive mappings and their examples

Let T be a mapping with domain D(T) and range R(T) in a Hilbert space H. Recall that T is said to be

  1. (i)

    pseudocontractive if

    TxTy,xy x y 2 ,x,yD(T),

    or, equivalently,

    T x T y 2 x y 2 + ( I T ) x ( I T ) y 2 ,x,yD(T);
  2. (ii)

    demicontractive if, for all xD(T) and pF(T),

    Txp,xp x p 2

or, equivalently,

T x p 2 x p 2 + ( I T ) x 2 ;
  1. (iii)

    k-demicontractive if there exists a constant k[0,1) such that

    T x p 2 x p 2 +k ( I T ) x 2 for all xD(T) and pF(T);
  2. (iv)

    quasi-nonexpansive if it is 0-demicontractive, that is,

    Txpxpfor all xD(T) and pF(T);
  3. (v)

    Lipschitzian if there exists L>0 such that

    TxTyLxy,x,yD(T);
  4. (vi)

    nonexpansive if it is Lipschitzian with L=1;

  5. (vii)

    contractive if it is Lipschitzian with L<1.

A Banach space (X,) is said to satisfy Opial’s condition if, for each sequence { x n } in X which converges weakly to a point xX, we have

lim inf n x n x< lim inf n x n y,yX,yx.

It is well known that any Hilbert space satisfies Opial’s condition.

Definition 2.1 (see [2])

Let K be a nonempty closed convex subset of a real Hilbert space H and T be a mapping from K into K. The mapping T is said to be demiclosed if, for any sequence { x n } which weakly converges to y, and if the sequence {T x n } strongly converges to z, then Ty=z.

Remark 2.1 In Definition 2.1, the particular case of demiclosedness at zero is frequently used in some iterative convergence algorithms, which is the particular case when z=θ, the zero vector of H; for more details, one can refer to [2].

The following concept of zero-demiclosedness was introduced as follows.

Definition 2.2 (see [[25], Definition 2.3])

Let K be a nonempty closed convex subset of a real Hilbert space and T be a mapping from K into K. The mapping T is called zero-demiclosed if { x n } in K satisfying x n T x n 0 and x n zK implies Tz=z.

The following result was essentially proved in [25], but we give the proof for the sake of completeness and the reader’s convenience.

Theorem 2.1 (see [[25], Proposition 2.4])

Let K be a nonempty closed convex subset of a real Hilbert space with zero vector θ. Then the following statements hold.

  1. (a)

    Let T be a mapping from K into K. Then T is zero-demiclosed if and only if IT is demiclosed at θ.

  2. (b)

    Let T be a nonexpansive mapping from H into itself. If there is a bounded sequence { x n }H such that x n T x n 0 as n0, then T is zero-demiclosed.

Proof Obviously, the conclusion (a) holds. To see (b), since { x n } is bounded, there is a subsequence { x n k }{ x n } and zH such that x n k z. One can claim Tz=z. Indeed, if Tzz, it follows from Opial’s condition that

lim inf k x n k z < lim inf k x n k T z lim inf k { x n k T x n k + T x n k T z } = lim inf k T x n k T z lim inf k x n k z ,

which is a contradiction. So Tz=z and hence T is zero-demiclosed. □

Now, we give some examples to show the existence of these generalized contractive mappings (i)-(vi) which also expound the relation between them.

Example A Let H=R with the absolute-value norm || and C=[2,0]. Let T:CC be defined by

Tx= { x 2 2 if  x [ 1 , 0 ] , 1 if  x [ 2 , 1 ] .

Then F(T)={1}. Since

|Tx(1) | 2 |x(1) | 2 + 1 2 | T x x | 2 for all xC,

we know that T is a 1 2 -demicontractive mapping. However, due to

|T ( 1 2 ) (1)|>| 1 2 (1)|,

T is not quasi-nonexpansive.

Example B Let H=R with the absolute-value norm || and C=[ 1 2 ,2]. Let T:CC be defined by

Tx= 1 x ,xC.

Then F(T)={1}. Since

| T x 1 | 2 | x 1 | 2 + 3 4 | T x x | 2 for all xC,

T is a 3 4 -demicontractive mapping. Moreover, T is also a pseudocontractive mapping.

Example C Let H=R with the absolute-value norm ||. Let T:HH be defined by

Tx= { ( 1 + x ) if  x 2 , x + 1 if  x 2 .

It is easy to see that

|TxTy||xy|for all x,yH.

So T is continuous nonexpansive with F(T)=.

The following example shows that there exists a continuous quasi-nonexpansive mapping which is not nonexpansive.

Example D (see [8])

Let H=R with the absolute-value norm || and C=[0,+). Define T:CC by

Tx= x 2 + 2 1 + x ,xC.

Obviously, F(T)={2}. It is easy to see that

|Tx2|= x 1 + x |x2||x2|for all xC

and

|T(0)T ( 1 3 ) |= 5 12 >|0 1 3 |.

Hence T is a continuous quasi-nonexpansive mapping but not nonexpansive.

The following example shows that there exists a demicontractive mapping which is neither pseudocontractive nor k-demicontractive for all k[0,1).

Example E Let H=R with the absolute-value norm ||. Let T:HH be defined by

Tx= { x 2 x + 1 if  x ( , 1 ] , x 2 + 1 1 + x if  x [ 1 , + ) .

Then F(T)={1}. Since

| T x 1 | 2 | x 1 | 2 + | T x x | 2 for all xH,

T is a demicontractive mapping. However, T is not a pseudocontractive mapping due to the fact that when x=3 and y=2.5, we have

| T x T y | 2 > | x y | 2 +|(xTx)(yTy) | 2 .

It is easy to see that T is not a k-demicontractive mapping for all k[0,1).

The following example shows that there exists a discontinuous pseudocontractive mapping which is not a demicontractive mapping.

Example F Let H=R with the absolute-value norm ||. Let T:HH be defined by

Tx= { x 2 + 1 if  x ( , 0 ] , 1 x 2 if  x ( 0 , + ) .

Then F(T)=. Due to

| T x T y | 2 | x y | 2 +|(IT)x(IT)y | 2 for all xH,

we know that T is a discontinuous pseudocontractive mapping but not a demicontractive mapping.

The following example shows that there exists a pseudocontractive mapping which is not k-demicontractive for all k[0,1).

Example G Let H=R with the absolute-value norm ||. Let T:HH be defined by

Tx= { 2 x 2 if  x [ 0 , 1 ] , 2 x if  x [ 1 , 2 ] , 0 if  x [ 1 , + ) .

Then F(T)={1}. Since

| T x T y | 2 | x y | 2 +|(IT)x(IT)y | 2 for all xH,

T is a pseudocontractive mapping. It is easy to see that T is not a k-demicontractive mapping for all k[0,1).

The following example shows that there exists a discontinuous k-demicontractive mapping for some k[0,1) as well as being demiclosed at θ which is neither pseudocontractive nor quasi-nonexpansive.

Example H Let H=R with the absolute-value norm || and C=[2,0]. Let T:CC be defined by

Tx= { x 2 2 if  x [ 1 , 0 ] , 1 8 if  x = 3 2 , 1 if  x [ 2 , 3 2 ) ( 3 2 , 1 ] .

Then the following statements hold.

  1. (a)

    T is discontinuous 3 4 -demicontractive.

  2. (b)

    T is demiclosed at θ.

  3. (c)

    T is not pseudocontractive.

  4. (d)

    T is not quasi-nonexpansive.

Proof Clearly, F(T)={1}. Since

|Tx(1) | 2 |x(1) | 2 + 3 4 |(IT)x | 2 for all xC,

T is a discontinuous 3 4 -demicontractive mapping and (a) is proved. Now, we verify (b). In fact, let { x n }[2,0] with x n z and x n T x n 0 as n. If all x n [1,0], we can prove Tz=z and z=1F(T) easily. If there exists a subsequence { x n k }[2,1], then, from x n T x n 0 as n, we can find a subsequence { x n k i } of { x n k } such that x n k i 3 2 for all i. Hence we have

|z(1)||z x n k i |+| x n k i T x n k i |+|T x n k i (1)|0as i,

which implies z=1F(T). To see (c) and (d), note that

|T ( 3 2 ) T ( 25 16 ) | 2 >| 3 2 ( 25 16 ) | 2 +|(IT) ( 3 2 ) (IT) ( 25 16 ) | 2

and

|T ( 3 2 ) (1)|>| ( 3 2 ) (1)|,

so T is neither pseudocontractive nor quasi-nonexpansive. The proof is completed. □

3 New feasible iterative algorithms for SCSP and strong convergence theorems

In this section, we establish some new strong convergence theorems by using feasible iterative algorithms for SCSP.

Theorem 3.1 Let H 1 and H 2 be two real Hilbert spaces and θ i be the zero vector of H i for i=1,2. Let C be a nonempty closed convex subset of H 1 and A: H 1 H 2 be a bounded linear operator with its adjoint B. Let T:CC be a Lipschitzian pseudocontractive mapping with Lipschitz constant L>0 and F(T), and let S: H 2 H 2 be a k-demicontractive mapping with F(S) which is demiclosed at θ 2 . Let C 1 =C and { x n } be a sequence generated by the following algorithm:

{ x 1 C 1 chosen arbitrarily , y n = ( 1 α ) x n + α T x n , z n = β x n + ( 1 β ) T y n , w n = P C ( z n + ξ B ( S I ) A z n ) , C n + 1 = { v C n : w n v z n v x n v } , x n + 1 = P C n + 1 ( x 1 ) , n N ,
(3.1)

where 0<1β<α< 1 2 1 + L 2 , ξ(0, 1 k B 2 ) and P C n is the projection operator from H 1 into C n for nN. Suppose that

Ω= { p F ( T ) : A p F ( S ) } .

Then there exists qΩ such that

  1. (a)

    x n q as n,

  2. (b)

    A x n Aq as n.

Proof We will show the conclusion by proceeding with the following steps.

Step 1. For any pΩ, we prove

w n p 2 z n p 2 ξ ( 1 k ξ B 2 ) ( S I ) A z n 2 .
(3.2)

Indeed, since

w n p 2 z n + ξ B ( S I ) A z n p 2 = z n p 2 + ξ B ( S I ) A z n 2 + 2 ξ z n p , B ( S I ) A z n = z n p 2 + ξ B ( S I ) A z n 2 + 2 ξ A z n A p , ( S I ) A z n = z n p 2 + ξ B ( S I ) A z n 2 + 2 ξ A z n A p + ( S I ) A z n ( S I ) A z n , ( S I ) A z n = z n p 2 + ξ B ( S I ) A z n 2 + 2 ξ S A z n A p , ( S I ) A z n 2 ξ ( S I ) A z n 2 z n p 2 + ξ 2 B 2 ( S I ) A z n 2 + 2 ξ S A z n A p , ( S I ) A z n 2 ξ ( S I ) A z n 2

and

2 ξ S A z n A p , ( S I ) A z n = ξ { S A z n A p 2 + ( S I ) A z n 2 A z n A p 2 } ξ { A z n A p 2 + k ( S I ) A z n 2 + ( S I ) A z n 2 A z n A p 2 } ξ { A z n A z n 2 + k ( S I ) A z n 2 + ( S I ) A z n 2 } = ξ { k ( S I ) A z n 2 + ( S I ) A z n 2 } ,

we get

w n p 2 z n p 2 ξ ( 1 k ξ B 2 ) ( S I ) A z n 2 ,

and our desired result is proved.

Step 2. We prove

z n p x n pfor all nN.
(3.3)

For any nN, by (3.1), we have

z n p 2 = β x n p 2 + ( 1 β ) T y n p 2 ( 1 β ) β T y n x n 2 β x n p 2 + ( 1 β ) y n p 2 + ( 1 β ) T y n y n 2 ( 1 β ) β T y n x n 2 β x n p 2 + ( 1 β ) ( ( 1 α ) x n p 2 + α T x n x n 2 ( 1 α ) α T x n x n 2 ) + ( 1 β ) T y n y n 2 ( 1 β ) β T y n x n 2 x n p 2 + ( 1 β ) ( α T x n x n 2 ( 1 α ) α T x n x n 2 ) ( 1 β ) β T y n x n 2 + ( 1 β ) ( 1 α ) ( x n T y n ) + α ( T x n T y n ) 2 x n p 2 + ( 1 β ) ( α T x n x n 2 ( 1 α ) α T x n x n 2 ) ( 1 β ) β T y n x n 2 + ( 1 β ) ( ( 1 α ) x n T y n 2 + α T x n T y n 2 ( 1 α ) α T x n x n 2 ) x n p 2 + ( 1 β ) ( α T x n x n 2 ( 1 α ) α T x n x n 2 ) ( 1 β ) β T y n x n 2 + ( 1 β ) ( ( 1 α ) x n T y n 2 + α L 2 x n y n 2 ( 1 α ) α T x n x n 2 ) x n p 2 + ( 1 β ) ( α T x n x n 2 ( 1 α ) α T x n x n 2 ) ( 1 β ) β T y n x n 2 + ( 1 β ) ( ( 1 α ) x n T y n 2 + α 3 L 2 x n T x n 2 ( 1 α ) α T x n x n 2 ) = x n p 2 ( 1 β ) ( α + β 1 ) T y n x n 2 ( 1 β ) α ( 1 2 α α 2 L 2 ) T x n x n 2 .
(3.4)

Since α+β>1 and α< 1 2 1 + L 2 , from (3.4), we have z n p 2 x n p 2 , or, equivalently,

z n p x n p.
(3.5)

Step 3. We show that C n is a nonempty closed convex set for any nN.

For any pΩ, by taking into account (3.2) and (3.5), we obtain

w n p z n p x n pfor all nN.

So we know Ω C n and hence C n for all nN. It is easy to verify that C n is closed and convex for all nN.

Step 4. We prove that { x n } is a Cauchy sequence in C and x n q as n for some qC.

Since Ω C n + 1 C n and x n + 1 = P C n + 1 ( x 1 ) C n , we get

x n + 1 x 1 p x 1 for all pΩ

and

x n x 1 x n + 1 x 1 for all nN,

which show that { x n } is bounded and { x n x 1 } is nondecreasing in [0,). So

lim n x n x 1 0

exists. For any m,nN with m>n, from x m = P C m ( x 1 ) C n and (1.1), we have

x m x n 2 + x 1 x n 2 = x m P C n ( x 1 ) 2 + x 1 P C n ( x 1 ) 2 x m x 1 2 .
(3.6)

Inequality (3.6) implies

lim m , n x n x m =0.

So { x n } is a Cauchy sequence. Clearly,

lim n x n + 1 x n =0.
(3.7)

By the completeness of C, there exists qC such that x n q as n.

Step 5. Finally, we show that the following hold:

  1. (i)

    qΩ,

  2. (ii)

    A x n Aq as n.

For any nN, since x n + 1 = P C n + 1 ( x 1 ) C n + 1 C n , from (3.1), we have

z n x n z n x n + 1 + x n + 1 x n 2 x n + 1 x n
(3.8)

and

w n x n w n x n + 1 + x n + 1 x n 2 x n + 1 x n .
(3.9)

From inequalities (3.7), (3.8) and (3.9), we deduce

lim n z n x n = 0 , lim n w n x n = 0
(3.10)

and hence

lim n w n z n =0.
(3.11)

By taking into account (3.4) and (3.10), we get

α ( 1 2 α α 2 L 2 ) T x n x n 2 + ( α + β 1 ) T y n x n 2 1 1 β ( x n p 2 z n p 2 ) 2 1 β x n z n x n p 0 as  n .

So, we obtain

lim n T x n x n = lim n T y n x n =0.
(3.12)

Since x n q as n, from (3.12) and the continuity of the norm and the Lipschitzian pseudocontractive mapping T, we can deduce that Tq=q, namely qF(T). On the other hand, from (3.2) and (3.11), we have

ξ ( 1 k ξ B 2 ) ( S I ) A z n 2 z n p 2 w n p 2 z n w n ( z n p w n p ) 0 as  n ,

which yields that

lim n ( S I ) A z n =0.
(3.13)

Since the k-demicontractive mapping S is demiclosed at θ 2 , taking into account x n q, A x n Aq, z n x n 0 and (3.13), we have

A z n Aq

and

AqF(S).

Hence we confirm qΩ. The proof is completed. □

By virtue of Theorem 3.1, we can establish the following:

  1. (i)

    Strong convergence algorithms for the split common solution problem for Lipschitzian pseudocontractive mappings and nonexpansive mappings (see Corollary 3.1 below).

  2. (ii)

    Strong convergence algorithms for the split common solution problem for Lipschitzian pseudocontractive mappings and quasi-nonexpansive mappings (see Corollary 3.2 below).

Corollary 3.1 Let H 1 and H 2 be two real Hilbert spaces and θ i be the zero vector of H i for i=1,2. Let C be a nonempty closed convex subset of H 1 and A: H 1 H 2 be a bounded linear operator with its adjoint B. Let T:CC be a Lipschitzian pseudocontractive mapping with Lipschitz constant L>0 and F(T), and let S: H 2 H 2 be a nonexpansive mapping with F(S). Let C 1 =C and { x n } be a sequence generated by the following algorithm:

{ x 1 C 1 chosen arbitrarily , y n = ( 1 α ) x n + α T x n , z n = β x n + ( 1 β ) T y n , w n = P C ( z n + ξ B ( S I ) A z n ) , C n + 1 = { v C n : w n v z n v x n v } , x n + 1 = P C n + 1 ( x 1 ) , n N ,

where 0<1β<α< 1 2 1 + L 2 , ξ(0, 1 B 2 ) and P C n is the projection operator from H 1 into C n for nN. Suppose that

Ω= { p F ( T ) : A p F ( S ) } .

Then there exists qΩ such that

  1. (a)

    x n q as n,

  2. (b)

    A x n Aq as n.

Proof Since the mapping S is nonexpansive, it is 0-demicontractive. Hence the desired conclusion follows from Theorem 3.1 immediately by taking k=0. □

Corollary 3.2 Let H 1 and H 2 be two real Hilbert spaces and θ i be the zero vector of H i for i=1,2. Let C be a nonempty closed convex subset of H 1 and A: H 1 H 2 be a bounded linear operator with its adjoint B. Let T:CC be a Lipschitzian pseudocontractive mapping with Lipschitz constant L>0 and F(T), and let S: H 2 H 2 be a quasi-nonexpansive mapping with F(S) which is demiclosed at θ 2 . Let C 1 =C and { x n } be a sequence generated by the following algorithm:

{ x 1 C 1 chosen arbitrarily , y n = ( 1 α ) x n + α T x n , z n = β x n + ( 1 β ) T y n , w n = P C ( z n + ξ B ( S I ) A z n ) , C n + 1 = { v C n : w n v z n v x n v } , x n + 1 = P C n + 1 ( x 1 ) , n N ,

where 0<1β<α< 1 2 1 + L 2 , ξ(0, 1 B 2 ) and P C n is the projection operator from H 1 into C n for nN. Suppose that

Ω= { p F ( T ) : A p F ( S ) } .

Then there exists qΩ such that

  1. (a)

    x n q as n,

  2. (b)

    A x n Aq as n.

Example 3.1 Let H 1 =R with the absolute-value norm ||. Let H 2 = [ 1 2 , 2 ] 2 with the norm α= ( a 1 2 + a 2 2 ) 1 2 for α=( a 1 , a 2 ) H 2 and the inner product α,β= i = 1 2 a i b i for α=( a 1 , a 2 ) and β=( b 1 , b 2 ) H 2 . Let A: H 1 H 2 be defined by Ax=(x,x) for xR. Then A is a bounded linear operator with its adjoint operator Bz= z 1 + z 2 for z=( z 1 , z 2 ) H 2 . Clearly, A=B= 2 . Let C=[ 1 2 , 2 ]. Let T:CC and S: H 2 H 2 be defined by

Tx= 1 x for xC

and

Sz= ( 1 z 1 , 1 z 2 ) for z=( z 1 , z 2 ) H 2 ,

respectively. It is easy to see that

  • F(T)={1};

  • F(S)={(1,1)};

  • Ω={pF(T):ApF(S)}={1};

  • T is a Lipschitzian pseudocontractive mapping with Lipschitz constant L= 2 ;

  • T and S both are 3 4 -demicontractive mappings.

By using algorithm (3.1) with 0<1β<α< 1 2 3 and ξ(0, 1 8 ), we can verify x n 1 and A x n A(1)=(1,1)F(S) as n.

4 Some applications and further remarks for Theorem 3.1

Let C be a nonempty subset of a Hilbert space H. Recall that a mapping U:CC is said to be accretive if

UxUy,xy0for all x,yC.

Obviously, U:CC is accretive if and only if IU:CC is pseudocontractive. Moreover,

F(IU)= U 1 (θ):={xC:Ux=θ},

where θ is the zero vector of H.

At the end of this paper, by applying Theorem 3.1, we obtain the following:

  1. (i)

    Strong convergence algorithms for the split common solution problem for Lipschitzian accretive mappings and demicontractive nonexpansive mappings (see Theorem 4.1 below).

  2. (ii)

    Strong convergence algorithms for the split common solution problem for Lipschitzian accretive mappings and nonexpansive mappings (see Corollary 4.1 below).

  3. (iii)

    Strong convergence algorithms for the split common solution problem for Lipschitzian accretive mappings and quasi-nonexpansive mappings (see Corollary 4.2 below).

Theorem 4.1 Let H 1 and H 2 be two real Hilbert spaces and θ i be the zero vector of H i for i=1,2. Let A: H 1 H 2 be a bounded linear operator with its adjoint B and U: H 1 H 1 be a Lipschitzian accretive mapping with Lipschitz constant L>0 and U 1 ( θ 1 ). Let S: H 2 H 2 be a k-demicontractive mapping with F(S) which is demiclosed at θ 2 . Let { x n } be a sequence generated by the following algorithm:

{ x 1 H 1 chosen arbitrarily , y n = x n α U x n , z n = β x n + ( 1 β ) ( I U ) y n , w n = z n + ξ B ( S I ) A z n , C n + 1 = { v C n : w n v z n v x n v } , x n + 1 = P C n + 1 ( x 1 ) , n N ,
(4.1)

where 0<1β<α< 1 2 1 + L 2 , ξ(0, 1 k B 2 ) and P C n is the projection operator from H 1 into C n for nN. Suppose that

Ω= { p U 1 ( θ 1 ) : A p F ( S ) } .

Then there exists qΩ such that

  1. (a)

    x n q as n,

  2. (b)

    A x n Aq as n.

Proof Let C 1 = H 1 . Then the iterative process (4.1) can be rewritten as follows:

{ x 1 C 1 chosen arbitrarily , y n = ( 1 α ) x n + α ( I U ) x n , z n = β x n + ( 1 β ) ( I U ) y n , w n = z n + ξ B ( S I ) A z n , C n + 1 = { v C n : w n v z n v x n v } , x n + 1 = P C n + 1 ( x 1 ) , n N .

Set T:=IU, then F(T)= U 1 ( θ 1 ) and T is a Lipschitzian pseudocontractive mapping with Lipschitz constant 1+L. Therefore the desired conclusion follows from Theorem 3.1 immediately. □

The following interesting results are immediate from Theorem 4.1.

Corollary 4.1 Let H 1 and H 2 be two real Hilbert spaces and θ i be the zero vector of H i for i=1,2. Let A: H 1 H 2 be a bounded linear operator with its adjoint B and U: H 1 H 1 be a Lipschitzian accretive mapping with Lipschitz constant L>0 and U 1 ( θ 1 ). Let S: H 2 H 2 be a quasi-nonexpansive mapping with F(S) which is demiclosed at θ 2 . Let { x n } be a sequence generated by the following algorithm:

{ x 1 H 1 chosen arbitrarily , y n = x n α U x n , z n = β x n + ( 1 β ) ( I U ) y n , w n = z n + ξ B ( S I ) A z n , C n + 1 = { v C n : w n v z n v x n v } , x n + 1 = P C n + 1 ( x 1 ) , n N ,

where 0<1β<α< 1 2 1 + L 2 , ξ(0, 1 B 2 ) and P C n is the projection operator from H 1 into C n for nN. Suppose that

Ω= { p U 1 ( θ 1 ) : A p F ( S ) } .

Then there exists qΩ such that

  1. (a)

    x n q as n,

  2. (b)

    A x n Aq as n.

Corollary 4.2 Let H 1 and H 2 be two real Hilbert spaces and θ i be the zero vector of H i for i=1,2. Let A: H 1 H 2 be a bounded linear operator with its adjoint B. Let U: H 1 H 1 be a Lipschitzian accretive mapping with Lipschitz constant L>0 and U 1 ( θ 1 ). Let S: H 2 H 2 be a nonexpansive mapping with F(S). Let { x n } be a sequence generated by the following algorithm:

{ x 1 H 1 chosen arbitrarily , y n = x n α U x n , z n = β x n + ( 1 β ) ( I U ) y n , w n = z n + ξ B ( S I ) A z n , C n + 1 = { v C n : w n v z n v x n v } , x n + 1 = P C n + 1 ( x 1 ) , n N ,

where 0<1β<α< 1 2 1 + L 2 , ξ(0, 1 B 2 ) and P C n is the projection operator from H 1 into C n for nN. Suppose that

Ω= { p U 1 ( θ 1 ) : A p F ( S ) } .

Then there exists qΩ such that

  1. (a)

    x n q as n,

  2. (b)

    A x n Aq as n.

Remark 4.1 In Theorems 3.1 and 4.1, the control coefficients α and β can be respectively replaced with the sequences { α n } and { β n } satisfying 0<ε<1 β n < α n < 1 2 1 + L 2 for some positive real number ε.

Remark 4.2 Obviously, all results in this paper are true if H 1 = H 2 . They generalize and improve many results in the literature; see, for instance, [23, 24, 2629].

References

  1. Censor Y, Segal A: The split common fixed point problem for directed operators. J. Convex Anal. 2009, 16: 587–600.

    MathSciNet  Google Scholar 

  2. Moudafi A: A note on the split common fixed-point problem for quasi-nonexpansive operators. Nonlinear Anal. 2011, 74: 4083–4087. 10.1016/j.na.2011.03.041

    Article  MathSciNet  Google Scholar 

  3. Moudafi A: Split monotone variational inclusions. J. Optim. Theory Appl. 2011, 150: 275–283. 10.1007/s10957-011-9814-6

    Article  MathSciNet  Google Scholar 

  4. Yao Y, Postolache M, Liou Y-C: Strong convergence of a self-adaptive method for the split feasibility problem. Fixed Point Theory Appl. 2013., 2013: Article ID 201

    Google Scholar 

  5. Censor Y, Gibali A, Reich S: Algorithms for the split variational inequality problem. Numer. Algorithms 2012, 59(2):301–323. 10.1007/s11075-011-9490-5

    Article  MathSciNet  Google Scholar 

  6. Zhao J, He S: Strong convergence of the viscosity approximation process for the split common fixed-point problem of quasi-nonexpansive mapping. J. Appl. Math. 2012., 2012: Article ID 438023 10.1155/2012/438023

    Google Scholar 

  7. He Z: The split equilibrium problems and its convergence algorithms. J. Inequal. Appl. 2012., 2012: Article ID 162

    Google Scholar 

  8. He Z, Du W-S: Nonlinear algorithms approach to split common solution problems. Fixed Point Theory Appl. 2012., 2012: Article ID 130

    Google Scholar 

  9. He Z, Du W-S: On hybrid split problem and its nonlinear algorithms. Fixed Point Theory Appl. 2013., 2013: Article ID 47

    Google Scholar 

  10. Li C-l, Liou Y-C, Yao Y: A damped algorithm for the split feasibility and fixed point problems. J. Inequal. Appl. 2013., 2013: Article ID 379

    Google Scholar 

  11. Moudafi A: The split common fixed-point problem for demicontractive mappings. Inverse Probl. 2010., 26: Article ID 055007

    Google Scholar 

  12. Ceng L-C, Petrusel A, Yao J-C: Strong convergence of modified implicit iterative algorithms with perturbed mappings for continuous pseudocontractive mappings. Appl. Math. Comput. 2009, 209: 162–176. 10.1016/j.amc.2008.10.062

    Article  MathSciNet  Google Scholar 

  13. Chen R, Song Y, Zhou H: Convergence theorems for implicit iteration process for a finite family of continuous pseudocontractive mappings. J. Math. Anal. Appl. 2006, 314: 701–709. 10.1016/j.jmaa.2005.04.018

    Article  MathSciNet  Google Scholar 

  14. Chidume CO, Souza GD: Convergence of a Halpern-type iteration algorithm for a class of pseudo-contractive mappings. Nonlinear Anal. 2008, 69: 2286–2292. 10.1016/j.na.2007.08.008

    Article  MathSciNet  Google Scholar 

  15. Morales CH, Jung JS: Convergence of paths for pseudocontractive mappings in Banach spaces. Proc. Am. Math. Soc. 2000, 128: 3411–3419. 10.1090/S0002-9939-00-05573-8

    Article  MathSciNet  Google Scholar 

  16. Yao Y, Liou Y-C, Chen R: Strong convergence of an iterative algorithm for pseudocontractive mappings in Banach spaces. Nonlinear Anal. 2007, 67: 3311–3317. 10.1016/j.na.2006.10.013

    Article  MathSciNet  Google Scholar 

  17. Zhou H: Strong convergence of an explicit iterative algorithm for continuous pseudo-contractions in Banach spaces. Nonlinear Anal. 2009, 70: 4039–4046. 10.1016/j.na.2008.08.012

    Article  MathSciNet  Google Scholar 

  18. Schu J: Approximating fixed points of Lipschitzian pseudocontractive mappings. Houst. J. Math. 1993, 19: 107–115.

    MathSciNet  Google Scholar 

  19. Udomene A: Path convergence, approximation of fixed points and variational solutions of Lipschitz pseudocontractions in Banach spaces. Nonlinear Anal. 2007, 67: 2403–2414. 10.1016/j.na.2006.09.001

    Article  MathSciNet  Google Scholar 

  20. Song Y: Strong convergence of viscosity approximation methods with strong pseudocontraction for Lipschitz pseudocontractive mappings. Positivity 2009, 13: 643–655. 10.1007/s11117-008-2246-3

    Article  MathSciNet  Google Scholar 

  21. Song Y: A note on the paper ‘A new iteration process for generalized Lipschitz pseudocontractive and generalized Lipschitz accretive mappings’. Nonlinear Anal. 2008, 68: 3047–3049. 10.1016/j.na.2007.02.044

    Article  MathSciNet  Google Scholar 

  22. Ishikawa S: Fixed point by a new iteration method. Proc. Am. Math. Soc. 1974, 4(1):147–150.

    Article  MathSciNet  Google Scholar 

  23. He Z: A new iterative scheme for equilibrium problems and fixed point problems of strict pseudo-contractive mappings and its application. Math. Commun. 2012, 17: 411–422.

    MathSciNet  Google Scholar 

  24. Yao Y, Postolache M, Liou Y-C: Coupling Ishikawa algorithms with hybrid techniques for pseudocontractive mappings. Fixed Point Theory Appl. 2013., 2013: Article ID 211

    Google Scholar 

  25. Du, W-S, He, Z: Feasible iterative algorithms for split common solution problems. J. Nonlinear Convex Anal. (in press)

  26. Zhou H: Convergence theorems of fixed points for Lipschitz pseudo-contractions in Hilbert spaces. J. Math. Anal. Appl. 2008, 343: 546–556. 10.1016/j.jmaa.2008.01.045

    Article  MathSciNet  Google Scholar 

  27. Tang YC, Peng JG, Liu LW: Strong convergence theorem for pseudocontractive mappings in Hilbert spaces. Nonlinear Anal. 2011, 74: 380–385. 10.1016/j.na.2010.08.048

    Article  MathSciNet  Google Scholar 

  28. Chidume CE, Zegeye H: Approximate fixed point sequences and convergence theorems for Lipschitz pseudo-contractive maps. Proc. Am. Math. Soc. 2003, 132: 831–840.

    Article  MathSciNet  Google Scholar 

  29. Shahzad N, Zegeye H: Approximating a common point of fixed points of a pseudocontractive mapping and zeros of sum of monotone mappings. Fixed Point Theory Appl. 2014., 2014: Article ID 85

    Google Scholar 

Download references

Acknowledgments

The first author was supported by the Candidate Foundation of Youth Academic Experts at Honghe University (2014HB0206); the second author was supported by Grant No. MOST 103-2115-M-017-001 of the Ministry of Science and Technology of the Republic of China.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Wei-Shih Du.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

Both authors contributed equally and significantly in writing this paper. Both authors read and approved the final manuscript.

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

Verify currency and authenticity via CrossMark

Cite this article

He, Z., Du, WS. On split common solution problems: new nonlinear feasible algorithms, strong convergence results and their applications. Fixed Point Theory Appl 2014, 219 (2014). https://doi.org/10.1186/1687-1812-2014-219

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/1687-1812-2014-219

Keywords

  • Lipschitzian
  • demicontractive mapping
  • pseudocontractive mapping
  • quasi-nonexpansive mapping
  • nonexpansive mapping
  • split common solution problem (SCSP)
  • iterative algorithm
  • strong convergence theorem