Open Access

Viscosity approximation methods for nonexpansive mappings in CAT(0) spaces

Journal of Inequalities and Applications20132013:93

https://doi.org/10.1186/1029-242X-2013-93

Received: 28 November 2012

Accepted: 9 February 2013

Published: 6 March 2013

Abstract

The purpose of this paper is to study the strong convergence theorems of Moudafi’s viscosity approximation methods for a nonexpansive mapping T in CAT(0) spaces without the property . For a contraction f on C and t ( 0 , 1 ) , let x t C be the unique fixed point of the contraction x t f ( x ) ( 1 t ) T x ; i.e.,

x t = t f ( x t ) ( 1 t ) T x t

and

x n + 1 = α n f ( x n ) ( 1 α n ) T x n , n 0 ,

where x 0 C is arbitrarily chosen and { α n } ( 0 , 1 ) satisfies certain conditions. We prove that the iterative schemes { x t } and { x n } converge strongly to the same point x ˜ such that x ˜ = P F ( T ) f ( x ˜ ) , which is the unique solution of the variational inequality (VIP)

x ˜ f x ˜ , x x ˜ 0 , x F ( T ) .

By using the concept of quasilinearization, we remark that the proof is different from that of Shi and Chen in J. Appl. Math. 2012:421050, 2012. In fact, strong convergence theorems for two given iterative schemes are established in CAT(0) spaces without the property .

Keywords

viscosity approximation method nonexpansive mapping variational inequality CAT(0) space common fixed point

1 Introduction

Let ( X , d ) be a metric space. A geodesic path joining x X to y X (or, more briefly, a geodesic from x to y) is a map c from a closed interval [ 0 , l ] R to X such that c ( 0 ) = x , c ( l ) = y , and d ( c ( t ) , c ( t ) ) = | t t | for all t , t [ 0 , l ] . In particular, c is an isometry and d ( x , y ) = l . The image α of c is called a geodesic (or metric) segment joining x and y. When it is unique, this geodesic segment is denoted by [ x , y ] . The space ( X , d ) is said to be a geodesic space if every two points of X are joined by a geodesic, and X is said to be uniquely geodesic if there is exactly one geodesic joining x and y for each x , y X . A subset Y X is said to be convex if Y includes every geodesic segment joining any two of its points. A geodesic triangle ( x 1 , x 2 , x 3 ) in a geodesic metric space ( X , d ) consists of three points x 1 , x 2 , x 3 in X (the vertices of ) and a geodesic segment between each pair of vertices (the edges of ). A comparison triangle for the geodesic triangle ( x 1 , x 2 , x 3 ) in ( X , d ) is a triangle ¯ ( x 1 , x 2 , x 3 ) : = ( x ¯ 1 , x ¯ 2 , x ¯ 3 ) in the Euclidean plane E 2 such that d E 2 ( x ¯ i , x ¯ j ) = d ( x i , x j ) for all i , j { 1 , 2 , 3 } .

A geodesic space is said to be a CAT(0) space if all geodesic triangles of an appropriate size satisfy the following comparison axiom.

CAT(0): Let be a geodesic triangle in X and let ¯ be a comparison triangle for . Then is said to satisfy the CAT(0) inequality if for all x , y and all comparison points x ¯ , y ¯ ¯ ,
d ( x , y ) d E 2 ( x ¯ , y ¯ ) .
If x, y 1 , y 2 are points in a CAT(0) space and if y 0 is the midpoint of the segment [ y 1 , y 2 ] , then the CAT(0) inequality implies
d 2 ( x , y 0 ) 1 2 d 2 ( x , y 1 ) + 1 2 d 2 ( x , y 2 ) 1 4 d 2 ( y 1 , y 2 ) .
(CN)

This is the (CN)-inequality of Bruhat and Tits [1]. In fact (cf. [2], p.163), a geodesic space is a CAT(0) space if and only if it satisfies the (CN)-inequality.

It is well known that any complete, simply connected Riemannian manifold having nonpositive sectional curvature is a CAT(0) space. Other examples include pre-Hilbert spaces, -trees (see [2]), Euclidean buildings (see [3]), the complex Hilbert ball with a hyperbolic metric (see [4]), and many others. Complete CAT(0) spaces are often called Hadamard spaces.

It is proved in [2] that a normed linear space satisfies the (CN)-inequality if and only if it satisfies the parallelogram identity, i.e., is a pre-Hilbert space; hence it is not so unusual to have an inner product-like notion in Hadamard spaces. Berg and Nikolaev [5] introduced the concept of quasilinearization as follows:

Let us formally denote a pair ( a , b ) X × X by a b and call it a vector. Then quasilinearization is defined as a map , : ( X × X ) × ( X × X ) R defined by
a b , c d = 1 2 ( d 2 ( a , d ) + d 2 ( b , c ) d 2 ( a , c ) d 2 ( b , d ) ) ( a , b , c , d X ) .
(1)
It is easily seen that a b , c d = c d , a b , a b , c d = b a , c d , and a x , c d + x b , c d = a b , c d for all a , b , c , d , x X . We say that X satisfies the Cauchy-Schwarz inequality if
a b , c d d ( a , b ) d ( c , d )
(2)

for all a , b , c , d X . It is known [[5], Corollary 3] that a geodesically connected metric space is a CAT(0) space if and only if it satisfies the Cauchy-Schwarz inequality.

In 2010, Kakavandi and Amini [6] introduced the concept of a dual space for CAT(0) spaces as follows. Consider the map Θ : R × X × X C ( X ) defined by
Θ ( t , a , b ) ( x ) = t a b , a x ,
(3)
where C ( X ) is the space of all continuous real-valued functions on X. Then the Cauchy-Schwarz inequality implies that Θ ( t , a , b ) is a Lipschitz function with the Lipschitz semi-norm L ( Θ ( t , a , b ) ) = | t | d ( a , b ) for all t R and a , b X , where
L ( f ) = sup { f ( x ) f ( y ) d ( x , y ) : x , y X , x y }
is the Lipschitz semi-norm of the function f : X R . Now, define the pseudometric D on R × X × X by
D ( ( t , a , b ) , ( s , c , d ) ) = L ( Θ ( t , a , b ) Θ ( s , c , d ) ) .

Lemma 1.1 [[6], Lemma 2.1]

D ( ( t , a , b ) , ( s , c , d ) ) = 0 if and only if t a b , x y = s c d , x y for all x , y X .

For a complete CAT(0) space ( X , d ) , the pseudometric space ( R × X × X , D ) can be considered as a subspace of the pseudometric space ( Lip ( X , R ) , L ) of all real-valued Lipschitz functions. Also, D defines an equivalence relation on R × X × X , where the equivalence class of t a b : = ( t , a , b ) is
[ t a b ] = { s c d : t a b , x y = s c d , x y x , y X } .

The set X : = { [ t a b ] : ( t , a , b ) R × X × X } is a metric space with a metric D, which is called the dual metric space of ( X , d ) .

Recently, Dehghan and Rooin [7] introduced the duality mapping in CAT(0) spaces and studied its relation with subdifferential by using the concept of quasilinearization. Then they presented a characterization of a metric projection in CAT(0) spaces as follows.

Theorem 1.2 [[7], Theorem 2.4]

Let C be a nonempty convex subset of a complete CAT(0) space X, x X and u C . Then
u = P C x if and only if y u , u x 0 for all y C .

Let C be a nonempty subset of a complete CAT(0) space X. Then the mapping T of C into itself is called nonexpansive iff d ( T x , T y ) d ( x , y ) for all x , y C . A point x C is called a fixed point of T if x = T x . We denote by F ( T ) the set of all fixed points of T. Kirk [8] showed that the fixed point set of a nonexpansive mapping T is closed and convex. A mapping f of C into itself is called contraction with coefficient α ( 0 , 1 ) iff d ( f x , f y ) α d ( x , y ) for all x , y C . Banach’s contraction principle [9] guarantees that f has a unique fixed point when C is a nonempty closed convex subset of a complete metric space. The existence of fixed points and convergence theorems for several mappings in CAT(0) spaces have been investigated by many authors (see also [1016]).

In 2010, Saejung [15] studied the convergence theorems of the following Halpern’s iterations for a nonexpansive mapping T: Let u be fixed and x t C be the unique fixed point of the contraction x t u ( 1 t ) T x ; i.e.,
x t = t u ( 1 t ) T x t ,
(4)
where t [ 0 , 1 ] and x 0 , u C are arbitrarily chosen and
x n + 1 = α n u ( 1 α n ) T x n , n 0 ,
(5)

where { α n } ( 0 , 1 ) . It is proved in [15] that { x t } converges strongly as t 0 to x ˜ F ( T ) which is nearest to u ( x ˜ = P F ( T ) u ), and { x n } converges strongly as n to x ˜ F ( T ) which is nearest to u under certain appropriate conditions on { α n } , where P C x is a metric projection from X onto C.

In 2012, Shi and Chen [16] studied the convergence theorems of the following Moudafi’s viscosity iterations for a nonexpansive mapping T: For a contraction f on C and t ( 0 , 1 ) , let x t C be the unique fixed point of the contraction x t f ( x ) ( 1 t ) T x ; i.e.,
x t = t f ( x t ) ( 1 t ) T x t ,
(6)
and x 0 C is arbitrarily chosen and
x n + 1 = α n f ( x n ) ( 1 α n ) T x n , n 0 ,
(7)
where { α n } ( 0 , 1 ) . They proved that { x t } defined by (6) converges strongly as t 0 to x ˜ F ( T ) such that x ˜ = P F ( T ) f ( x ˜ ) in the framework of a CAT(0) space satisfying the property , i.e., if for x , u , y 1 , y 2 X ,
d ( x , P [ x , y 1 ] u ) d ( x , y 1 ) d ( x , P [ x , y 2 ] u ) d ( x , y 2 ) + d ( x , u ) d ( y 1 , y 2 ) .

Furthermore, they also obtained that { x n } defined by (7) converges strongly as n to x ˜ F ( T ) under certain appropriate conditions imposed on { α n } .

All of the above bring us the following conjectures.

Question 1.3 Could we obtain the strong convergence of both { x t } and { x n } defined by (6) and (7) respectively, in the framework of a CAT(0) space without the property ?

The purpose of this paper is to study the strong convergence theorems of the iterative schemes (6) and (7) in CAT(0) spaces without the property . We prove that the iterative schemes (6) and (7) converge strongly to x ˜ such that x ˜ = P F ( T ) f ( x ˜ ) , which is the unique solution of the variational inequality (VIP)
x ˜ f x ˜ , x x ˜ 0 , x F ( T ) .
(8)

By using the concept of quasilinearization, we remark that the proof given below is different from that of Shi and Chen [16]. In fact, strong convergence theorems for two given iterative schemes are established in CAT(0) spaces without the property .

2 Preliminaries

In this paper, we write ( 1 t ) x t y for the unique point z in the geodesic segment joining from x to y such that
d ( z , x ) = t d ( x , y ) and d ( z , y ) = ( 1 t ) d ( x , y ) .

We also denote by [ x , y ] the geodesic segment joining from x to y, that is, [ x , y ] = { ( 1 t ) x t y : t [ 0 , 1 ] } . A subset C of a CAT(0) space is convex if [ x , y ] C for all x , y C .

The following lemmas play an important role in our paper.

Lemma 2.1 [[2], Proposition 2.2]

Let X be a CAT(0) space, p , q , r , s X and λ [ 0 , 1 ] . Then
d ( λ p ( 1 λ ) q , λ r ( 1 λ ) s ) λ d ( p , r ) + ( 1 λ ) d ( q , s ) .

Lemma 2.2 [[11], Lemma 2.4]

Let X be a CAT(0) space, x , y , z X and λ [ 0 , 1 ] . Then
d ( λ x ( 1 λ ) y , z ) λ d ( x , z ) + ( 1 λ ) d ( y , z ) .

Lemma 2.3 [[11], Lemma 2.5]

Let X be a CAT(0) space, x , y , z X and λ [ 0 , 1 ] . Then
d 2 ( λ x ( 1 λ ) y , z ) λ d 2 ( x , z ) + ( 1 λ ) d 2 ( y , z ) λ ( 1 λ ) d 2 ( x , y ) .

The concept of Δ-convergence introduced by Lim [17] in 1976 was shown by Kirk and Panyanak [18] in CAT(0) spaces to be very similar to weak convergence in the Banach space setting. Next, we give the concept of Δ-convergence and collect some basic properties.

Let { x n } be a bounded sequence in a CAT(0) space X. For x X , we set
r ( x , { x n } ) = lim sup n d ( x , x n ) .
The asymptotic radius r ( { x n } ) of { x n } is given by
r ( { x n } ) = inf { r ( x , { x n } ) : x X } ,
and the asymptotic center A ( { x n } ) of { x n } is the set
A ( { x n } ) = { x X : r ( x , { x n } ) = r ( { x n } ) } .
It is known from Proposition 7 of [14] that in a CAT(0) space, A ( { x n } ) consists of exactly one point. A sequence { x n } X is said to Δ-converge to x X if A ( { x n k } ) = { x } for every subsequence { x n k } of { x n } . The uniqueness of an asymptotic center implies that the CAT(0) space X satisfies Opial’s property, i.e., for given { x n } X such that { x n } Δ-converges to x and given y X with y x ,
lim sup n d ( x n , x ) < lim sup n d ( x n , y ) .

Since it is not possible to formulate the concept of demiclosedness in a CAT(0) setting, as stated in linear spaces, let us formally say that ‘ I T is demiclosed at zero’ if the conditions { x n } C Δ-converges to x and d ( x n , T x n ) 0 imply x F ( T ) .

Lemma 2.4 [18]

Every bounded sequence in a complete CAT(0) space always has a Δ-convergent subsequence.

Lemma 2.5 [13]

If C is a closed convex subset of a complete CAT(0) space and if { x n } is a bounded sequence in C, then the asymptotic center of { x n } is in C.

Lemma 2.6 [13]

If C is a closed convex subset of X and T : C X is a nonexpansive mapping, then the conditions { x n } Δ-converges to x and d ( x n , T x n ) 0 imply x C and T x = x .

Having the notion of quasilinearization, Kakavandi and Amini [6] introduced the following notion of convergence.

A sequence { x n } in the complete CAT(0) space ( X , d ) w-converges to x X if lim n x x n , x y = 0 , i.e., lim n ( d 2 ( x n , x ) d 2 ( x n , y ) + d 2 ( x , y ) ) = 0 for all y X .

It is obvious that convergence in the metric implies w-convergence, and it is easy to check that w-convergence implies Δ-convergence [[6], Proposition 2.5], but it is showed in ([[19], Example 4.7]) that the converse is not valid. However, the following lemma shows another characterization of Δ-convergence as well as, more explicitly, a relation between w-convergence and Δ-convergence.

Lemma 2.7 [[19], Theorem 2.6]

Let X be a complete CAT(0) space, { x n } be a sequence in X, and x X . Then { x n } Δ-converges to x if and only if lim sup n x x n , x y 0 for all y X .

Lemma 2.8 [[20], Lemma 2.1]

Let { a n } be a sequence of non-negative real numbers satisfying the property
a n + 1 ( 1 α n ) a n + α n β n , n 0 ,
where { α n } ( 0 , 1 ) and { β n } R such that
  1. 1.

    n = 0 α n = ;

     
  2. 2.

    lim sup n β n 0 or n = 0 | α n β n | < .

     

Then { a n } converges to zero as n .

The following two lemmas can be obtained from elementary computation. For convenience of the readers, we include the details.

Lemma 2.9 Let X be a complete CAT(0) space. Then for all u , x , y X , the following inequality holds:
d 2 ( x , u ) d 2 ( y , u ) + 2 x y , x u .
Proof
d 2 ( y , u ) d 2 ( x , u ) 2 y x , x u = d 2 ( y , u ) d 2 ( x , u ) 2 y u , x u 2 u x , x u = d 2 ( y , u ) d 2 ( x , u ) 2 y u , x u + 2 d 2 ( x , u ) = d 2 ( y , u ) + d 2 ( x , u ) 2 y u , x u d 2 ( y , u ) + d 2 ( x , u ) 2 d ( y , u ) d ( x , u ) = ( d ( y , u ) d ( x , u ) ) 2 0 .
Hence
d 2 ( x , u ) d 2 ( y , u ) + 2 x y , x u .

 □

Lemma 2.10 Let X be a CAT(0) space. For any u , v X and t [ 0 , 1 ] , let u t = t u ( 1 t ) v . Then, for all x , y X ,
  1. (i)

    u t x , u t y t u x , u t y + ( 1 t ) v x , u t y ;

     
  2. (ii)

    u t x , u y t u x , u y + ( 1 t ) v x , u y and u t x , v y t u x , v y + ( 1 t ) v x , v y .

     
Proof (i) It follows from the (CN)-inequality that
2 u t x , u t y = d 2 ( u t , y ) + d 2 ( x , u t ) d 2 ( u t , u t ) d 2 ( x , y ) t d 2 ( u , y ) + ( 1 t ) d 2 ( v , y ) t ( 1 t ) d 2 ( u , v ) + d 2 ( x , u t ) d 2 ( u t , u t ) d 2 ( x , y ) = t d 2 ( u , y ) + t d 2 ( x , u t ) t d 2 ( u , u t ) t d 2 ( x , y ) + ( 1 t ) d 2 ( v , y ) + ( 1 t ) d 2 ( x , u t ) ( 1 t ) d 2 ( v , u t ) ( 1 t ) d 2 ( x , y ) + t d 2 ( u , u t ) + ( 1 t ) d 2 ( v , u t ) t ( 1 t ) d 2 ( u , v ) = t [ d 2 ( u , y ) + d 2 ( x , u t ) d 2 ( u , u t ) d 2 ( x , y ) ] + ( 1 t ) [ d 2 ( v , y ) + d 2 ( x , u t ) d 2 ( v , u t ) d 2 ( x , y ) ] + t ( 1 t ) 2 d 2 ( u , v ) + ( 1 t ) t 2 d 2 ( u , v ) t ( 1 t ) d 2 ( u , v ) = t u x , u t y + ( 1 t ) v x , u t y .

(ii) The proof is similar to (i). □

3 Variational inequalities in CAT(0) spaces

In this section, we present strong convergence theorems of Moudafi’s viscosity methods in CAT(0) spaces. Our first result is the continuous version of Theorem 2.2 of Shi and Chen [16]. By using the concept of quasilinearization, we note that the proof given below is different from that of Shi and Chen.

For any t ( 0 , 1 ] and a contraction f with coefficient α ( 0 , 1 ) , define the mapping S t : C C by
G t = t f ( x ) ( 1 t ) T x , x C .
(9)
It is not hard to see that G t is a contraction on C. Indeed, for x , y C , we have
d ( G t ( x ) , G t ( y ) ) = d ( t f ( x ) ( 1 t ) T x , t f ( y ) ( 1 t ) T y ) d ( t f ( x ) ( 1 t ) T x , t f ( y ) ( 1 t ) T x ) + d ( t f ( y ) ( 1 t ) T x , t f ( y ) ( 1 t ) T y ) t d ( f ( x ) , f ( y ) ) + ( 1 t ) d ( T x , T y ) t α d ( x , y ) + ( 1 t ) d ( x , y ) = ( 1 t ( 1 α ) ) d ( x , y ) .
This implies that G t is a contraction mapping. Then there exists a unique u C such that
u = G t ( u ) = t f ( u ) ( 1 t ) T u .
Let x t C be the unique fixed point of G t . Thus
x t = t f ( x t ) ( 1 t ) T x t .
(10)
Theorem 3.1 Let C be a closed convex subset of a complete CAT(0) space X, and let T : C C be a nonexpansive mapping with F ( T ) . Let f be a contraction on C with coefficient 0 < α < 1 . For each t ( 0 , 1 ] , let { x t } be given by
x t = t f ( x t ) ( 1 t ) T x t .
(11)
Then { x t } converges strongly as t 0 to x ˜ such that x ˜ = P F ( T ) f ( x ˜ ) which is equivalent to the following variational inequality:
x ˜ f ( x ˜ ) , x x ˜ 0 , x F ( T ) .
(12)
Proof We first show that { x t } is bounded. For any p F ( T ) , we have that
d ( x t , p ) = d ( t f ( x t ) ( 1 t ) T x t , p ) t d ( f ( x t ) , p ) + ( 1 t ) d ( T x t , p ) t d ( f ( x t ) , p ) + ( 1 t ) d ( x t , p ) .
Then
d ( x t , p ) d ( f ( x t ) , p ) d ( f ( x t ) , f ( p ) ) + d ( f ( p ) , p ) α d ( x t , p ) + d ( f ( p ) , p ) .
This implies that
d ( x t , p ) 1 1 α d ( f ( p ) , p ) .
Hence { x t } is bounded, so are { T x t } and { f ( x t ) } . We get that
lim t 0 d ( x t , T x t ) = lim t 0 d ( t f ( x t ) ( 1 t ) T x t , T x t ) lim t 0 [ t d ( f ( x t ) , T x t ) + ( 1 t ) d ( T x t , T x t ) ] lim t 0 t d ( f ( x t ) , T x t ) = 0 .
Assume that { t n } ( 0 , 1 ) is such that t n 0  as  n . Put x n : = x t n . We will show that { x n } contains a subsequence converging strongly to x ˜ such that x ˜ = P F ( T ) f ( x ˜ ) which is equivalent to the following variational inequality:
x ˜ f ( x ˜ ) , x x ˜ 0 , x F ( T ) .
Since { x n } is bounded, by Lemma 2.4, 2.6, we may assume that { x n } Δ-converges to a point x ˜ and x ˜ F ( T ) . It follows from Lemma 2.10 (i) that
d 2 ( x n , x ˜ ) = x n x ˜ , x n x ˜ α n f ( x n ) x ˜ , x n x ˜ + ( 1 α n ) T x n x ˜ , x n x ˜ α n f ( x n ) x ˜ , x n x ˜ + ( 1 α n ) d ( T x n , x ˜ ) d ( x n , x ˜ ) α n f ( x n ) x ˜ , x n x ˜ + ( 1 α n ) d 2 ( x n , x ˜ ) .
It follows that
d 2 ( x n , x ˜ ) f ( x n ) x ˜ , x n x ˜ = f ( x n ) f ( x ˜ ) , x n x ˜ + f ( x ˜ ) x ˜ , x n x ˜ d ( f ( x n ) , f ( x ˜ ) ) d ( x n , x ˜ ) + f ( x ˜ ) x ˜ , x n x ˜ α d 2 ( x n , x ˜ ) + f ( x ˜ ) x ˜ , x n x ˜ ,
and thus
d 2 ( x n , x ˜ ) 1 1 α f ( x ˜ ) x ˜ , x n x ˜ .
(13)
Since { x n } Δ-converges to x ˜ , by Lemma 2.7, we have
lim sup n f ( x ˜ ) x ˜ , x n x ˜ 0 .
(14)

It follows from (13) that { x n } converges strongly to x ˜ .

Next, we show that x ˜ solves the variational inequality (12). Applying Lemma 2.3, for any q F ( T ) ,
d 2 ( x t , q ) = d 2 ( t f ( x t ) ( 1 t ) T x t , q ) t d 2 ( f ( x t ) , q ) + ( 1 t ) d 2 ( T x t , q ) t ( 1 t ) d 2 ( f ( x t ) , T x t ) t d 2 ( f ( x t ) , q ) + ( 1 t ) d 2 ( x t , q ) t ( 1 t ) d 2 ( f ( x t ) , T x t ) .
It implies that
d 2 ( x t , q ) d 2 ( f ( x t ) , q ) ( 1 t ) d 2 ( f ( x t ) , T x t ) .
Taking the limit through t = t n 0 , we can get that
d 2 ( x ˜ , q ) d 2 ( f ( x ˜ ) , q ) d 2 ( f ( x ˜ ) , x ˜ ) .
Hence
0 1 2 [ d 2 ( x ˜ , x ˜ ) + d 2 ( f ( x ˜ ) , q ) d 2 ( x ˜ , q ) d 2 ( f ( x ˜ ) , x ˜ ) ] = x ˜ f ( x ˜ ) , q x ˜ , q F ( T ) .

That is, x ˜ solves the inequality (12).

Finally, we show that the entire net { x t } converges to x ˜ , assume x s n x ˆ , where s n 0 . By the same argument, we get that x ˆ F ( T ) and solves the variational inequality (12), i.e.,
x ˜ f ( x ˜ ) , x ˜ x ˆ 0 ,
(15)
and
x ˆ f ( x ˆ ) , x ˆ x ˜ 0 .
(16)
Adding up (15) and (16), we get that
0 x ˜ f ( x ˜ ) , x ˜ x ˆ x ˆ f ( x ˆ ) , x ˜ x ˆ = x ˜ f ( x ˆ ) , x ˜ x ˆ + f ( x ˆ ) f ( x ˜ ) , x ˜ x ˆ x ˆ x ˜ , x ˜ x ˆ x ˜ f ( x ˆ ) , x ˜ x ˆ = x ˜ x ˆ , x ˜ x ˆ f ( x ˆ ) f ( x ˜ ) , x ˆ x ˜ x ˜ x ˆ , x ˜ x ˆ d ( f ( x ˆ ) , f ( x ˜ ) ) d ( x ˆ , x ˜ ) d 2 ( x ˜ , x ˆ ) α d ( x ˆ , x ˜ ) d ( x ˆ , x ˜ ) = d 2 ( x ˜ , x ˆ ) α d 2 ( x ˆ , x ˜ ) ( 1 α ) d 2 ( x ˜ , x ˆ ) .

Since 0 < α < 1 , we have that d ( x ˜ , x ˆ ) = 0 , and so x ˜ = x ˆ . Hence the net x t converges strongly to x ˜ which is the unique solution to the variational inequality (12). This completes the proof. □

Remark 3.2 We give the different proof of [[16], Theorem 2.2]. In fact, the property imposed on a CAT(0) space X is removed.

If f u , then the following result can be obtained directly from Theorem 3.1.

Corollary 3.3 [[15], Lemma 2.2]

Let C be a closed convex subset of a complete CAT(0) space X, and let T : C C be a nonexpansive mapping with F ( T ) . For each t ( 0 , 1 ] , let u be fixed and { x t } be given by
x t = t u ( 1 t ) T x t .
(17)
Then { x t } converges strongly as t 0 to x ˜ F ( T ) which is nearest to u which is equivalent to the following variational inequality:
x ˜ u , x x ˜ 0 , x F ( T ) .
(18)
Theorem 3.4 Let C be a closed convex subset of a complete CAT(0) space X, and let T : C C be a nonexpansive mapping with F ( T ) . Let f be a contraction on C with coefficient 0 < α < 1 . For the arbitrary initial point x 0 C , let { x n } be generated by
x n + 1 = α n f ( x n ) ( 1 α n ) T x n , n 0 ,
(19)
where { α n } ( 0 , 1 ) satisfies the following conditions:
  1. (i)

    lim n α n = 0 ;

     
  2. (ii)

    n = 0 α n = ;

     
  3. (iii)

    either n = 0 | α n + 1 α n | < or lim n ( α n + 1 / α n ) = 1 .

     

Then { x n } converges strongly as n to x ˜ such that x ˜ = P F ( T ) f ( x ˜ ) which is equivalent to the variational inequality (12).

Proof We first show that the sequence { x n } is bounded. For any p F ( T ) , we have that
d ( x n + 1 , p ) = d ( α n f ( x n ) ( 1 α n ) T x n , p ) α n d ( f ( x n ) , p ) + ( 1 α n ) d ( T x n , p ) α n ( d ( f ( x n ) , f ( p ) ) + d ( f ( p ) , p ) ) + ( 1 α n ) d ( T x n , p ) max { d ( x n , p ) , 1 1 α d ( f ( p ) , p ) } .
By induction, we have
d ( x n , p ) max { d ( x 0 , p ) , 1 1 α d ( f ( p ) , p ) } ,
for all n N . Hence { x n } is bounded, so are { T x n } and { f ( x n ) } . Next, we claim that lim n d ( x n + 1 , x n ) = 0 . To this end, we observe that
d ( x n + 1 , x n ) = d ( α n f ( x n ) ( 1 α n ) T x n , α n 1 f ( x n 1 ) ( 1 α n 1 ) T x n 1 ) d ( α n f ( x n ) ( 1 α n ) T x n , α n f ( x n ) ( 1 α n ) T x n 1 ) + d ( α n f ( x n ) ( 1 α n ) T x n 1 , α n f ( x n 1 ) ( 1 α n ) T x n 1 ) + d ( α n f ( x n 1 ) ( 1 α n ) T x n 1 , α n 1 f ( x n 1 ) ( 1 α n 1 ) T x n 1 ) ( 1 α n ) d ( T x n , T x n 1 ) + α n d ( f ( x n ) , f ( x n 1 ) ) + | α n α n 1 | d ( f ( x n 1 ) , T x n 1 ) ( 1 α n ) d ( x n , x n 1 ) + α n d ( f ( x n ) , f ( x n 1 ) ) + | α n α n 1 | d ( f ( x n 1 ) , T x n 1 ) ( 1 α n ) d ( x n , x n 1 ) + α n α d ( x n , x n 1 ) + | α n α n 1 | d ( f ( x n 1 ) , T x n 1 ) = ( 1 α n ( 1 α ) ) d ( x n , x n 1 ) + | α n α n 1 | d ( f ( x n 1 ) , T x n 1 ) .
By the conditions (ii) and (iii) and Lemma 2.8, we have
lim n d ( x n + 1 , x n ) = 0 .
(20)
It follows from (20) and condition (i) that
d ( x n , T x n ) d ( x n , x n + 1 ) + d ( x n + 1 , T x n ) = d ( x n , x n + 1 ) + d ( α n f ( x n ) ( 1 α n ) T x n , T x n ) d ( x n , x n + 1 ) + α n d ( f ( x n ) , T x n ) 0 as  n .
(21)
Let { x t } be a net in C such that
x t = t f ( x t ) ( 1 t ) T x t .
By Theorem 3.1, we have that { x t } converges strongly as t 0 to a fixed point x ˜ F ( T ) , which solves the variational inequality (12). Now, we claim that
lim sup n f ( x ˜ ) x ˜ , x n x ˜ 0 .
It follows from Lemma 2.10 (i) that
d 2 ( x t , x n ) = x t x n , x t x n t f ( x t ) x n , x t x n + ( 1 t ) T x t x n , x t x n = t f ( x t ) f ( x ˜ ) , x t x n + t f ( x ˜ ) x ˜ , x t x n + t x ˜ x t , x t x n + t x t x n , x t x n + ( 1 t ) T x t T x n , x t x n + ( 1 t ) T x n x n , x t x n t α d ( x t , x ˜ ) d ( x t , x n ) + t f ( x ˜ ) x ˜ , x t x n + t d ( x ˜ , x t ) d ( x t , x n ) + t d 2 ( x t , x n ) + ( 1 t ) d 2 ( x t , x n ) + ( 1 t ) d ( T x n , x n ) d ( x t , x n ) t α d ( x t , x ˜ ) M + t f ( x ˜ ) x ˜ , x t x n + t d ( x ˜ , x t ) M + t d 2 ( x t , x n ) + ( 1 t ) d 2 ( x t , x n ) + ( 1 t ) d ( T x n , x n ) M d 2 ( x t , x n ) + t α d ( x t , x ˜ ) M + t d ( x ˜ , x t ) M + d ( T x n , x n ) M + t f ( x ˜ ) x ˜ , x t x n ,
where M sup m , n 1 { d ( x t , x n ) } . This implies that
f ( x ˜ ) x ˜ , x n x t ( 1 + α ) d ( x t , x ˜ ) M + d ( T x n , x n ) t M .
(22)
Taking the limit as n first and then t 0 , the inequality (22) yields
lim sup t 0 lim sup n f ( x ˜ ) x ˜ , x n x t 0 .
Since x t x ˜ as t 0 and by the continuity of a metric distance d, we have, for any fixed n 0 ,
lim t 0 f ( x ˜ ) x ˜ , x n x t = lim t 0 1 2 [ d 2 ( f ( x ˜ ) , x t ) + d 2 ( x ˜ , x n ) d 2 ( f ( x ˜ ) , x n ) d 2 ( x ˜ , x t ) ] = 1 2 [ d 2 ( f ( x ˜ ) , x ˜ ) + d 2 ( x ˜ , x n ) d 2 ( f ( x ˜ ) , x n ) d 2 ( x ˜ , x ˜ ) ] = f ( x ˜ ) x ˜ , x n x ˜ .
It implies that, for any ε > 0 , there exists a δ > 0 such that
f ( x ˜ ) x ˜ , x n x ˜ < f ( x ˜ ) x ˜ , x n x t + ε , t ( 0 , δ ) .
(23)
Thus, by the upper limit as n first and then t 0 , the inequality in (23), we get that
lim sup n f ( x ˜ ) x ˜ , x n x ˜ ε .
Since ε is arbitrary, it follows that
lim sup n f ( x ˜ ) x ˜ , x n x ˜ 0 .
Finally, we prove that x n x ˜ as n . For any n N , we set y n = α n x ˜ ( 1 α n ) T x n . It follows from Lemma 2.9 and Lemma 2.10 (i), (ii) that
d 2 ( x n + 1 , x ˜ ) d 2 ( y n , x ˜ ) + 2 x n + 1 y n , x n + 1 x ˜ ( α n d ( x ˜ , x ˜ ) + ( 1 α n ) d ( T x n , x ˜ ) ) 2 + 2 [ α n f ( x n ) y n , x n + 1 x ˜ + ( 1 α n ) T x n y n , x n + 1 x ˜ ] ( 1 α n ) 2 d 2 ( x n , x ˜ ) + 2 [ α n α n f ( x n ) x ˜ , x n + 1 x ˜ + α n ( 1 α n ) f ( x n ) T x n , x n + 1 x ˜ + ( 1 α n ) α n T x n x ˜ , x n + 1 x ˜ + ( 1 α n ) ( 1 α n ) T x n T x n , x n + 1 x ˜ ] ( 1 α n ) 2 d 2 ( x n , x ˜ ) + 2 [ α n α n f ( x n ) x ˜ , x n + 1 x ˜ + α n ( 1 α n ) f ( x n ) T x n , x n + 1 x ˜ + ( 1 α n ) α n T x n x ˜ , x n + 1 x ˜ + ( 1 α n ) 2 d ( T x n , T x n ) d ( x n + 1 x ˜ ) ] = ( 1 α n ) 2 d 2 ( x n , x ˜ ) + 2 [ α n 2 f ( x n ) x ˜ , x n + 1 x ˜ + α n ( 1 α n ) f ( x n ) x ˜ , x n + 1 x ˜ ] = ( 1 α n ) 2 d 2 ( x n , x ˜ ) + 2 α n f ( x n ) x ˜ , x n + 1 x ˜ = ( 1 α n ) 2 d 2 ( x n , x ˜ ) + 2 α n f ( x n ) f ( x ˜ ) , x n + 1 x ˜ + 2 α n f ( x ˜ ) x ˜ , x n + 1 x ˜ ( 1 α n ) 2 d 2 ( x n , x ˜ ) + 2 α n α d ( x n , x ˜ ) d ( x n + 1 , x ˜ ) + 2 α n f ( x ˜ ) x ˜ , x n + 1 x ˜ ( 1 α n ) 2 d 2 ( x n , x ˜ ) + α n α ( d 2 ( x n , x ˜ ) + d 2 ( x n + 1 , x ˜ ) ) + 2 α n f ( x ˜ ) x ˜ , x n + 1 x ˜ ,
which implies that
d 2 ( x n + 1 , x ˜ ) 1 ( 2 α ) α n + α n 2 1 α α n d 2 ( x n , x ˜ ) + 2 α n 1 α α n f ( x ˜ ) x ˜ , x n + 1 x ˜ 1 ( 2 α ) α n 1 α α n d 2 ( x n , x ˜ ) + 2 α n 1 α α n f ( x ˜ ) x ˜ , x n + 1 x ˜ + α n 2 M ,
where M sup n 0 { d 2 ( x n , x ˜ ) } . It then follows that
d 2 ( x n + 1 , x ˜ ) ( 1 α n ) d 2 ( x n , x ˜ ) + α n β n ,
where
α n = 2 ( 1 α ) α n 1 α α n and β n = ( 1 α α n ) α n 2 ( 1 α ) M + 1 ( 1 α ) f ( x ˜ ) x ˜ , x n + 1 x ˜ .

Applying Lemma 2.8, we can conclude that x n x ˜ . This completes the proof. □

Remark 3.5 We give the different proof of [[16], Theorem 2.3]. In fact, the property imposed on a CAT(0) space X is removed.

If f u , then the following corollary can be obtained directly from Theorem 3.4.

Corollary 3.6 [[15], Theorem 2.3]

Let C be a closed convex subset of a complete CAT(0) space X, and let T : C C be a nonexpansive mapping with F ( T ) . Let u , x 0 C be arbitrarily chosen and { x n } be generated by
x n + 1 = α n u ( 1 α n ) T x n , n 0 ,
(24)
where { α n } ( 0 , 1 ) satisfies the following conditions:
  1. (i)

    lim n α n = 0 ;

     
  2. (ii)

    n = 0 α n = ;

     
  3. (iii)

    either n = 0 | α n + 1 α n | < or lim n ( α n + 1 / α n ) = 1 .

     

Then { x n } converges strongly as n to x ˜ F ( T ) which is nearest to u which is equivalent to the following variational inequality (18).

Declarations

Acknowledgements

The first author is supported by Naresuan university.

Authors’ Affiliations

(1)
Department of Mathematics, Faculty of Science, Naresuan University

References

  1. Bruhat F, Tits J: Groupes ré ductifs sur un corps local. I. Donêes radicielles valuées. Inst. Hautes Êtudes Sci. Publ. Math. 1972, 41: 5–251. 10.1007/BF02715544MathSciNetView ArticleGoogle Scholar
  2. Bridson M, Haefliger A: Metric Spaces of Nonpositive Curvature. Springer, Berlin; 1999.View ArticleGoogle Scholar
  3. Brown KS: Buildings. Springer, New York; 1989.View ArticleGoogle Scholar
  4. Goebel K, Reich S Series of Monographs and Textbooks in Pure and Applied Mathematics 83. In Uniform Convexity, Hyperbolic Geometry, and Nonexpansive Mappings. Dekker, New York; 1984.Google Scholar
  5. Berg ID, Nikolaev IG: Quasilinearization and curvature of Alexandrov spaces. Geom. Dedic. 2008, 133: 195–218. 10.1007/s10711-008-9243-3MathSciNetView ArticleGoogle Scholar
  6. Kakavandi BA, Amini M: Duality and subdifferential for convex functions on complete CAT(0) metric spaces. Nonlinear Anal. 2010, 73: 3450–3455. 10.1016/j.na.2010.07.033MathSciNetView ArticleGoogle Scholar
  7. Dehghan H, Rooin J: A characterization of metric projection in CAT(0) spaces. International Conference on Functional Equation 2012, 41–43.Google Scholar
  8. Kirk WA: Geodesic geometry and fixed point theory. Colecc. Abierta 64. In Seminar of Mathematical Analysis (Malaga/Seville, 2002/2003). University of Seville, Secretary Publication, Seville; 2003:195–225.Google Scholar
  9. Banach S: Sur les oṕration dans les ensembles abstraits et leur applications aux équations intégrales. Fundam. Math. 1922, 3: 133–181.Google Scholar
  10. Cho YJ, Ciric LB, Wang S: Convergence theorems for nonexpansive semigroups in CAT(0) spaces. Nonlinear Anal. 2011, 74: 6050–6059. 10.1016/j.na.2011.05.082MathSciNetView ArticleGoogle Scholar
  11. Dhompongsa S, Panyanak B: On Δ-convergence theorems in CAT(0) spaces. Comput. Math. Appl. 2008, 56: 2572–2579. 10.1016/j.camwa.2008.05.036MathSciNetView ArticleGoogle Scholar
  12. Dhompongsa S, Kaewkhao A, Panyanak B: Lim’s theorems for multivalued mappings in CAT(0) spaces. J. Math. Anal. Appl. 2005, 312: 478–487. 10.1016/j.jmaa.2005.03.055MathSciNetView ArticleGoogle Scholar
  13. Dhompongsa S, Kirk WA, Panyanak B: Nonexpansive set-valued mappings in metric and Banach spaces. J. Nonlinear Convex Anal. 2007, 8: 35–45.MathSciNetGoogle Scholar
  14. Dhompongsa S, Kirk WA, Sims B: Fixed points of uniformly Lipschitzian mappings. Nonlinear Anal. 2006, 65: 762–772. 10.1016/j.na.2005.09.044MathSciNetView ArticleGoogle Scholar
  15. Saejung S: Halpern’s iteration in CAT(0) spaces. Fixed Point Theory Appl. 2010., 2010: Article ID 471781 10.1155/2010/471781Google Scholar
  16. Shi LY, Chen RD: Strong convergence of viscosity approximation methods for nonexpansive mappings in CAT(0) spaces. J. Appl. Math. 2012., 2012: Article ID 421050 10.1155/2012/421050Google Scholar
  17. Lim TC: Remarks on some fixed point theorems. Proc. Am. Math. Soc. 1976, 60: 179–182. 10.1090/S0002-9939-1976-0423139-XView ArticleGoogle Scholar
  18. Kirk WA, Panyanak B: A concept of convergence in geodesic spaces. Nonlinear Anal. 2008, 68: 3689–3696. 10.1016/j.na.2007.04.011MathSciNetView ArticleGoogle Scholar
  19. Kakavandi BA: Weak topologies in complete CAT(0) metric spaces. Proc. Am. Math. Soc. 2013, 141(3):1029–1039. 10.1090/S0002-9939-2012-11743-5View ArticleGoogle Scholar
  20. Xu HK: An iterative approach to quadratic optimization. J. Optim. Theory Appl. 2003, 116: 659–678. 10.1023/A:1023073621589MathSciNetView ArticleGoogle Scholar

Copyright

© Wangkeeree and Preechasilp; licensee Springer 2013

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/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.