Open Access

Convergence results of multi-valued nonexpansive mappings in Banach spaces

Journal of Inequalities and Applications20142014:483

https://doi.org/10.1186/1029-242X-2014-483

Received: 30 April 2014

Accepted: 19 November 2014

Published: 4 December 2014

Abstract

The purpose of this paper is to establish two new iteration schemes as follows: x n = α n x n 1 + ( 1 α n ) y n , y n T x n , n 1 , x n = β n u + α n x n 1 + ( 1 α n β n ) y n , y n T x n , n 1 , for a multi-valued nonexpansive mapping T in a uniformly convex Banach space and prove that { x n } and { x n } converge strongly to a fixed point of T under some suitable conditions, respectively. Moreover, a gap in Sahu (Nonlinear Anal. 37:401-407, 1999) is found and revised.

MSC:47H04, 47H10, 47J25.

Keywords

multi-valued nonexpansive mappings fixed point strong convergence implicit iteration

1 Introduction

Let E be a Banach space, K be a nonempty, closed, and convex subset of E. We denote the family of all subsets of E by 2 E , the family of nonempty, closed, and bounded subsets of E by C B ( E ) , the family of nonempty compact subsets of E by C ( E ) , and the family of nonempty, compact, and convex subsets of E by C C ( E ) . Let H ( , ) be Hausdorff metric on C B ( E ) , i.e.,
H ( A , B ) = max { sup x A d ( x , B ) , sup x B d ( x , A ) } A , B C B ( E ) ,
(1.1)

where d ( x , B ) = inf { x y : y B } .

A multi-valued mapping T : K C B ( K ) is called nonexpansive (respectively, contractive), if for any x , y K , we have
H ( T x , T y ) x y ( respectively , H ( T x , T y ) k x y  for some  k [ 0 , 1 ) ) .
(1.2)

A point x is called a fixed point of T if x T x . In this paper, F ( T ) stands for the fixed point set of the mapping T.

In 1968, Markin [1] firstly established the nonexpansive multi-valued convergence results in Hilbert space. Banach’s contraction principle was extended to a multi-valued contraction by Nadler [2] in 1969.

In 1974, a breakthrough was achieved by TC Lim using Edelstein’s method of asymptotic centers [3].

Theorem 1.1 (Lim [4])

Let K be a nonempty, bounded, closed, and convex subset of a uniformly convex Banach space E and T : K C ( K ) a multi-valued nonexpansive mapping. Then T has a fixed point.

In 1990, Kirk and Massa [5] obtained another important result for the multi-valued nonexpansive mappings.

Theorem 1.2 (Kirk and Massa [5])

Let K be a nonempty, bounded, closed, and convex subset of a Banach space E and T : K C C ( K ) a multi-valued nonexpansive mapping. Suppose that the asymptotic center in K of each bounded sequence of E is nonempty and compact. Then T has a fixed point.

In 1999, Sahu [6] obtained the strong convergence theorems of the nonexpansive type and non-self multi-valued mappings for the following iteration process:
x n = t n x 0 + ( 1 t n ) u n , n 0 ,
(1.3)
where u n T x n , x 0 K , t n ( 0 , 1 ) , and lim n t n = 0 . Unfortunately, a gap exists in the proof of Theorem 1 in p.405 of [6]; there are the following inequalities:
x n u n , J ( x n v ) = x n v + v u n , J ( x n v ) x n v 2 u n v x n v x n v 2 x n v 2 = 0 .
(1.4)

Clearly, if the above inequality holds, then the inequality u n v x n v must be assumed, for all v F ( T ) , but this inequality does not hold generally, based on the definition of the Hausdorff metric H ( , ) on C B ( E ) , for all v T ( v ) .

Remark 1.1 To revise the gap we have found in Theorem 1 of [6], in this paper, we change the fixed point set of T ( F ( T ) : = { x K , x T x } ) into F ( T ) : = { x K , x = T x } . The above problem is solved easily. Indeed,
x n u n , J ( x n v ) = x n v + v u n , J ( x n v ) x n v 2 u n v x n v x n v 2 H ( T x n , T v ) x n v x n v 2 x n v 2 = 0 .
(1.5)

In 2001, Xu [7] extended Theorem 1.2 to a multi-valued nonexpansive nonself-mapping and obtained the fixed theorems. Some recent fixed point results for multi-valued nonexpansive mappings can be found in [813] and the references therein.

Motivated by Sahu [6] and the above results, we propose two new iteration processes (1.6) and (1.7) and we study them in this paper. Let K be a nonempty, bounded, closed, and convex subset of E, and u and x 0 be fixed elements of K. We have
x n = α n x n 1 + ( 1 α n ) y n , y n T x n , n 1 ,
(1.6)
x n = β n u + α n x n 1 + ( 1 α n β n ) y n , y n T x n , n 1 ,
(1.7)

where α n , β n ( 0 , 1 ) satisfy certain conditions, and we prove some strongly convergence theorems for the multi-valued nonexpansive mappings in Banach spaces. The results presented in this paper mainly extend and improve the corresponding results of Sahu [6] on the iteration algorithms.

2 Preliminaries

Let E be a real uniformly convex Banach space and let J denote the normalized duality mapping from E to 2 E defined by
J ( x ) = { f E , x , f = x f , x = f } , x E ,
(2.1)

where E denotes the dual space of E and , denotes the generalized duality pair. It is well known that if E is smooth or if E is strictly convex, then J is single-valued.

Recall that the norm of Banach space E is said to be Gâteaux differentiable (or E is said to be smooth), if the limit
lim t 0 x + t y x t
(2.2)

exists for each x, y on the unit sphere S ( E ) of E. Moreover, if for each y in S ( E ) the limit defined by (2.2) is uniformly attained for x in S ( E ) , we say that the norm of E is uniformly Gâteaux differentiable. It is also well known that if E has a uniformly Gâteaux differentiable norm, then the duality mapping J is norm-to-weak star uniformly continuous on each bounded subset of E.

A Banach space E is called uniformly convex, if for each ϵ > 0 there is a δ > 0 such that for x , y E with x , y 1 , and x y ϵ , x + y 2 ( 1 δ ) holds. The modulus of convexity of E is defined by
δ E ( ϵ ) = inf { 1 1 2 ( x + y ) : x , y 1 , x y ϵ } ,

for all ϵ [ 0 , 2 ] . E is said to be uniformly convex if δ E ( 0 ) = 0 , and δ ( ϵ ) > 0 for all 0 < ϵ 2 .

Throughout this paper, we write x n x (respectively, x n x ) to indicate that the sequence x n weakly (respectively, weak*) converges to x, and as usual x n x will symbolize strong convergence. In order to show our main results, the following definitions and lemmas are needed.

Let LIM be a continuous linear functional on l satisfying LIM = 1 = LIM ( 1 ) . Then we know that LIM is a mean on N if and only if
inf { a n ; n N } LIM ( a ) sup { a n ; n N }
for every a = ( a 1 , a 2 , ) l . According to time and circumstances, we use LIM n ( a n ) instead of LIM ( a ) . A mean LIM on N is called a Banach limit if
LIM n ( a n ) = LIM n ( a n + 1 )

for every a = ( a 1 , a 2 , ) l . Furthermore, we have the following results [14, 15].

Lemma 2.1 ([[14], Lemma 1])

Let C be a nonempty, closed, and convex subset of a Banach space E with uniformly Gâteaux differentiable norm. Let { x n } be a bounded sequence of E and let LIM n be a mean LIM on N and z C . Then
LIM n x n z 2 = min LIM n x n y 2
if and only if
LIM n y z , J ( x n z ) 0

for all y C .

Definition 2.1 A multi-valued mapping T : K C B ( K ) is said to satisfy Condition I if there is a nondecreasing function f : [ 0 , ) [ 0 , ) with f ( 0 ) = 0 , f ( r ) > 0 for r ( 0 , ) such that
d ( x , T x ) f ( d ( x , F ( T ) ) ) for all  x K .

An example of mappings that satisfy Condition I can be found in reference [16].

3 Strong convergence theorems

Theorem 3.1 Let E be a uniformly convex Banach space with a uniformly Gâteaux differentiable norm, K be a nonempty closed convex subset of E, T : K C ( K ) be a multi-valued nonexpansive mapping. Assume that F ( T ) and T ( y ) = { y } for each y F ( T ) . Let { x n } be an implicit Mann type iteration defined by (1.6), where α n ( 0 , 1 ) and lim n α n = 0 , then the sequence { x n } converges strongly to a fixed point of T.

Proof Firstly, let p F ( T ) , n 1 , we show that lim n x n p exists and { x n } is bounded. Using (2.1), we obtain
x n p 2 = α n x n 1 + ( 1 α n ) y n p , j ( x n p ) = ( 1 α n ) y n p , j ( x n p ) + α n x n 1 p , j ( x n p ) ( 1 α n ) y n p x n p + α n x n 1 p x n p ( 1 α n ) H ( T x n , T p ) x n p + α n x n 1 p x n p ( 1 α n ) x n p 2 + α n x n 1 p x n p ,
so
x n p 2 x n 1 p x n p .
(3.1)

If x n p = 0 , then lim n x n p = 0 apparently holds.

If x n p > 0 , from (3.1) we have
x n p x n 1 p ,
we find that { x n p } is a decreasing sequence, so
lim n x n p

exists. Hence { x n } is bounded, and so is { y n } .

Secondly, we show that lim n x n y n = 0 .

It follows from (1.6) and from { x n } and y n T x n being bounded, that there exists a real number M 1 > 0 such that
x n 1 y n M 1 , x n y n = α n x n 1 y n α n M 1 .
Since lim n α n = 0 , we have
x n y n 0 , as  n ,
therefore
lim n x n y n = 0 .
(3.2)
We define ϕ : K [ 0 , ) by ϕ ( x ) : = LIM n x n x 2 for each x K , since E is uniformly convex (hence reflexive) and ϕ is continuous, convex, and ϕ ( x ) as x , ϕ attains its infimum over K (see [17]). Let
M : = { z K : ϕ ( z ) = min x K ϕ ( x ) } .

Then M is nonempty, bounded, closed, and convex (see [[15], Theorem 1.3.11]).

Next, we show that M is singleton.

Since M and { x n } are bounded, there exists R > 0 such that M, { x n } B R ( 0 ) for all n 0 . Then, by inequality (3.12) of [18] for y 1 , y 2 M we have
x n 1 2 ( y 1 + y 2 ) 2 1 2 x n y 1 2 + 1 2 x n y 2 2 1 4 g ( y 1 y 2 ) .
If y 1 y 2 , we have
r : = ϕ ( 1 2 ( y 1 + y 2 ) ) 1 2 ϕ ( y 1 ) + 1 2 ϕ ( y 2 ) 1 4 g ( y 1 y 2 ) < r .

This is a contradiction. Therefore, M has a unique element.

Now we show that z M is the fixed point of T. Since T is compact valued, we have some w n T z for z M such that
y n w n = d ( y n , T z ) H ( T x n , T z ) x n z ;

indeed, w n = P T z ( y n ) , since Tz is a compact set, and we take lim n w n = w T z .

It follows from the above inequality that
LIM n x n w LIM n x n y n + LIM n y n w n + LIM n w n w = LIM n y n w n LIM n x n z ,

and hence w M , it follows that M is singleton, so w = z T z .

Therefore, z F ( T ) and so T z = { z } by the assumption, and F ( T ) is nonempty.

On the other hand, for v F ( T ) , we have
x n y n , J ( x n v ) = x n v + v y n , J ( x n v ) = x n v 2 y n v , J ( x n v ) x n v 2 y n v x n v x n v 2 H ( T x n , T v ) x n v x n v 2 x n v x n v = 0 .
It follows from (1.6) that
0 x n y n , J ( x n v ) = α n x n 1 y n , J ( x n v )
(3.3)
and
x n x n 1 , J ( x n v ) = x n y n , J ( x n v ) + y n x n 1 , J ( x n v ) x n y n x n v + y n x n 1 , J ( x n v ) .
(3.4)
Hence, from (3.2), (3.3), and (3.4), we obtain
LIM n x n x n 1 , J ( x n v ) = 0 , v F ( T ) .
(3.5)
Finally, we show that LIM n x n 1 v , J ( x n v ) 0 . Let s ( 0 , 1 ) , then by Lemma 1 of [19] we get
x n z s ( x n 1 z ) 2 x n z 2 + 2 s ( x n 1 z ) , j ( x n z s ( x n 1 z ) ) = x n z 2 2 s x n 1 z , j ( x n z ) 2 s x n 1 z , j ( x n z s ( x n 1 z ) ) j ( x n z ) .
Let ε > 0 be arbitrary, then since J is norm-to-weak uniformly continuous on bounded subsets of E, there exists δ > 0 such that for all s ( 0 , δ ) , we have
LIM n x n 1 z , j ( x n z ) 1 2 s [ LIM n x n z 2 LIM n x n z s ( x n 1 z ) 2 ] + ε < ε ,
since z M , and it is a minimizer of ϕ over K. Now, since ε is arbitrary this implies that
LIM n x n 1 z , j ( x n z ) 0 .
(3.6)
Combining inequalities (3.5) and (3.6) we get
LIM n x n z , j ( x n z ) = LIM n x n z 2 0 .

Therefore, there is a subsequence { x n j } of { x n } which converges strongly to z.

Since lim n x n p exists p F ( T ) , we get
lim n x n z = 0 .

The proof is completed. □

Theorem 3.2 Let E be a uniformly convex Banach space with a uniformly Gâteaux differentiable norm, K a nonempty, closed, and convex subset of E, T : K C B ( K ) be a multi-valued nonexpansive mapping that satisfies Condition I, assume F ( T ) and T ( y ) = { y } for each y F ( T ) . Let { x n } be an implicit Mann type iteration defined by (1.6), where α n ( 0 , 1 ) and lim n α n = 0 , then the sequence { x n } converges strongly to a fixed point of T.

Proof It follows from the proof of Theorem 3.1 that
lim n d ( x n , T x n ) = 0 .
Then Condition I implies that
lim n d ( x n , F ( T ) ) = 0 .

The remainder of the proof is the same as Theorem 2.4 of [20]. □

Theorem 3.3 Let E be a uniformly convex Banach space with a uniformly Gâteaux differentiable norm, K be a nonempty, closed, and convex subset of E, T : K C ( K ) be a multi-valued nonexpansive mapping. Assume that F ( T ) and T ( y ) = { y } for each y F ( T ) . Let { x n } be the modified implicit Mann type iteration defined by (1.7), where α n , β n ( 0 , 1 ) and lim n α n β n = lim n β n = 0 . Then the sequence { x n } converges strongly to a fixed point of T.

Proof Firstly, let λ n = β n 1 α n and z n = λ n u + ( 1 λ n ) y n , then
x n = α n x n 1 + ( 1 α n ) z n .
(3.7)
Let p F ( T ) , n 1 ; we show that { x n } is bounded. Using (2.1), we obtain
x n p 2 = α n x n 1 + ( 1 α n ) z n p , j ( x n p ) = ( 1 α n ) z n p , j ( x n p ) + α n x n 1 p , j ( x n p ) ( 1 α n ) λ n ( u p ) , j ( x n p ) + ( 1 α n ) ( 1 λ n ) y n p , j ( x n p ) + α n x n 1 p x n p ( 1 α n ) ( λ n u p x n p + ( 1 λ n ) H ( T x n , T p ) x n p ) + α n x n 1 p x n p β n u p x n p + ( 1 α n β n ) x n p 2 + α n x n 1 p x n p ,
so
x n p 2 β n α n + β n u p x n p + α n α n + β n x n 1 p x n p .
(3.8)

If x n p = 0 , then { x n } is apparently bounded.

If x n p > 0 , from (3.7) we have
x n p β n α n + β n u p + α n α n + β n x n 1 p max { x n 1 p , u p } max { x 0 p , u p } .

Thus, { x n } is bounded.

Secondly, we show that lim n x n y n = 0 .

It follows from (1.7) and since { x n } and y n T x n are bounded, there exists a real number M 2 > 0 such that
x n 1 z n M 2 , x n z n = α n x n 1 z n α n M 2 .
Since lim n α n = 0 , we have
x n z n 0 as  n ,
therefore
lim n x n z n = 0 .
Since
x n y n x n z n + z n y n = x n z n + β n 1 α n u y n ,
then
lim n x n y n = 0 .
(3.9)
We define ϕ : K [ 0 , ) by ϕ ( x ) : = LIM n x n x 2 for each x K . Since E uniformly convex (hence reflexive) and ϕ is continuous, convex and ϕ ( x ) as x , ϕ attains its infimum over K (see, e.g., [17]). Let
M : = { z K : ϕ ( z ) = min x K ϕ ( x ) } .

Then M is nonempty, and it is also bounded, closed, and convex (see [[15], Theorem 1.3.11]).

In the same way as the proof of Theorem 3.1, we see that M is also singleton, and it follows from (3.9) that z is the fixed point of T.

By Lemma 2.1, we have
LIM n x z , J ( x n z ) 0 ,
for all x K . In particular, we have
LIM n u z , J ( x n z ) 0 .
(3.10)
It follows from the proof of (3.6) of Theorem 3.1 that (3.6) also holds here. Thus
x n z 2 = β n ( u z ) + α n ( x n 1 z ) + ( 1 α n β n ) ( y n z ) , J ( x n z ) β n u z , J ( x n z ) + α n x n 1 z , J ( x n z ) + ( 1 α n β n ) { H ( T x n , T z ) x n z } β n u z , J ( x n z ) + α n x n 1 z , J ( x n z ) + ( 1 α n β n ) x n z 2 .
Hence
x n z 2 β n α n + β n u z , J ( x n z ) + α n α n + β n x n 1 z , J ( x n z ) u z , J ( x n z ) + α n α n + β n ( x n 1 z , J ( x n z ) u z , J ( x n z ) ) u z , J ( x n z ) + α n α n + β n M 0 ,
where M 0 is a constant such that | x n 1 u , J ( x n z ) | M 0 . Thus
LIM n x n z 2 LIM n u z , J ( x n z ) + LIM n α n β n α n β n + 1 M 0 0 .
Therefore, there is a subsequence { x n j } of { x n } which converges strongly to z. To complete the proof, suppose there is another subsequence { x n k } of { x n } which converges strongly to  z ¯ . Then z ¯ is a fixed point of T by (3.9) and z ¯ K because K is closed. It then follows from (3.10) that
z ¯ z , J ( z ¯ z ) = z ¯ z 2 0 .

This proves the strong convergence of { x n } to z F ( T ) .

The proof is completed. □

4 Numerical examples

Now, we give two real numerical examples in which the conditions satisfy the ones of Theorem 3.1 and Theorem 3.3.

Example 4.1 Let E = R , K = [ 1 , 1 ] , T ( x ) = 1 5 x , x E , whichis nonexpansive, α n = 1 n for every n N . Then { x n } is the sequence generated by
x n = 1 n x n 1 + ( 1 1 n ) 1 5 x n ,
(4.1)

and x n 0 as n , where 0 is the fixed point of T.

Example 4.2 Let E = R , K = [ 1 , 1 ] , T ( x ) = 1 2 x , x E , which is nonexpansive, α n = 1 n 100 , β n = 1 n 50 for every n N . Then { x n } is the sequence generated by
x n = 1 n 50 + 1 n 100 x n 1 + ( 1 1 n 100 1 n 50 ) 1 2 x n ,
(4.2)

and x n 0 as n , where 0 is the fixed point of T.

Remark 4.1 We can prove Example 4.1 and Example 4.2 by Theorem 3.1 and Theorem 3.3, respectively, and we show two numerical experiments (Figure 1 and Figure 2) which can explain that the sequence { x n } strongly converges to 0.
Figure 1

x ( 1 ) = 1 , iteration steps n = 100 .

Figure 2

x ( 1 ) = 0.2 , iteration steps n = 1 , 000 .

Remark 4.2 From the above numerical examples, we can see that the convergence results in this paper are important. The main reason is that the convergence of the two iteration schemes in this paper can easily be implemented by the software of Matlab 7.0, so they can be applied for numerical calculations in practical problems.

Declarations

Acknowledgements

This work is supported by the Fundamental Research Funds for the Central Universities (No. K5051370004), the Natural Science Basic Research Plan in Shaanxi Province of China (No. 2014JQ1020), the National Science Foundation of China (No. 61202178 and No. 61373174), the National Natural Science Foundation of China, Tian Yuan Special Foundation (No. 11426167).

Authors’ Affiliations

(1)
School of Mathematics and Statistics, Xidian University
(2)
Department of Mathematics, Tianjin Polytechnic University

References

  1. Markin JT: A fixed point theorem for set valued mappings. Bull. Am. Math. Soc. 1968, 74: 639–640. 10.1090/S0002-9904-1968-11971-8MathSciNetView ArticleMATHGoogle Scholar
  2. Nadler SB: Multi-valued contraction mappings. Pac. J. Math. 1969, 30: 475–488. 10.2140/pjm.1969.30.475MathSciNetView ArticleMATHGoogle Scholar
  3. Edelstein M: The construction of an asymptotic center with a fixed point property. Bull. Am. Math. Soc. 1972, 78: 206–208. 10.1090/S0002-9904-1972-12918-5MathSciNetView ArticleMATHGoogle Scholar
  4. Lim TC: A fixed point theorem for multi-valued nonexpansive mappings in a uniformly convex Banach space. Bull. Am. Math. Soc. 1974, 80: 1123–1126. 10.1090/S0002-9904-1974-13640-2View ArticleMATHGoogle Scholar
  5. Kirk WA, Massa S: Remarks on asymptotic and Chebyshev centers. Houst. J. Math. 1990, 16: 357–364.MathSciNetMATHGoogle Scholar
  6. Sahu DR: Strong convergence theorems for nonexpansive type and non-self multi-valued mappings. Nonlinear Anal. 1999, 37: 401–407. 10.1016/S0362-546X(98)00056-XMathSciNetView ArticleMATHGoogle Scholar
  7. Xu HK: Multi-valued nonexpansive mappings in Banach spaces. Nonlinear Anal. 2001, 43: 693–706. 10.1016/S0362-546X(99)00227-8MathSciNetView ArticleMATHGoogle Scholar
  8. Reich S: A fixed point theorem for locally contractive multi-valued functions. Rev. Roum. Math. Pures Appl. 1972, 17: 569–572.MathSciNetMATHGoogle Scholar
  9. Feng Y, Liu S: Fixed point theorems for multi-valued contractive mappings and multi-valued Caristi type mappings. J. Math. Anal. Appl. 2006, 317: 103–112. 10.1016/j.jmaa.2005.12.004MathSciNetView ArticleMATHGoogle Scholar
  10. Benavides TD, Gavira B: The fixed point property for multi-valued nonexpansive mappings. J. Math. Anal. Appl. 2007, 328: 1471–1483. 10.1016/j.jmaa.2006.06.059MathSciNetView ArticleMATHGoogle Scholar
  11. Shahzad N, Zegeye H: On Mann and Ishikawa iteration schemes for multi-valued maps in Banach spaces. Nonlinear Anal. 2009, 71: 838–844. 10.1016/j.na.2008.10.112MathSciNetView ArticleMATHGoogle Scholar
  12. Jung JS: Strong convergence theorems for multi-valued nonexpansive nonself-mappings in Banach spaces. Nonlinear Anal. 2007, 66: 2345–2354. 10.1016/j.na.2006.03.023MathSciNetView ArticleMATHGoogle Scholar
  13. Alimohammady M, et al.: Iterative algorithms for a new class of extended general nonconvex set-valued variational inequalities. Nonlinear Anal. 2010, 73: 3907–3923. 10.1016/j.na.2010.08.022MathSciNetView ArticleMATHGoogle Scholar
  14. Takahashi W, Ueda Y: On Reich’s strong convergence for resolvents of accretive operators. J. Math. Anal. Appl. 1984, 104: 546–553. 10.1016/0022-247X(84)90019-2MathSciNetView ArticleMATHGoogle Scholar
  15. Takahashi W: Nonlinear Functional Analysis. Fixed Point Theory and Its Applications. Yokohama Publishers, Yokohama; 2000. (in Japanese)MATHGoogle Scholar
  16. Panyanak B: Mann and Ishikawa iterative processes for multi-valued mappings in Banach spaces. Comput. Math. Appl. 2007, 54: 872–877. 10.1016/j.camwa.2007.03.012MathSciNetView ArticleMATHGoogle Scholar
  17. Barbu V, Precupanu T: Convexity and Optimization in Banach Spaces. Editura Academiei RSR, Bucharest; 1978.MATHGoogle Scholar
  18. Xu HK: Inequalities in Banach spaces with applications. Nonlinear Anal. 1991, 16: 1127–1138. 10.1016/0362-546X(91)90200-KMathSciNetView ArticleMATHGoogle Scholar
  19. Morales CH, Jung JS: Convergence of paths for pseudocontractive mappings in Banach spaces. Proc. Am. Math. Soc. 2000, 128: 3411–3419. 10.1090/S0002-9939-00-05573-8MathSciNetView ArticleMATHGoogle Scholar
  20. Song Y: Convergence of iterative algorithms for multi-valued mappings in Banach spaces. Nonlinear Anal. 2009, 70: 1547–1556. 10.1016/j.na.2008.02.034MathSciNetView ArticleMATHGoogle Scholar

Copyright

© He et al.; licensee Springer. 2014

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.