Mann iteration for direct pseudocontractive maps

Abstract

For \(T\) a direct pseudocontractive map, we prove the convergence of Mann iteration to the fixed point of \(T\). In this note we introduce a new class of maps. Let \(X\) be a real normed space, and \(B\subset X\) be a nonemepty set. The map \(T:B\rightarrow B\) is direct pseudocontractive if there exists \(k\in \left( 0,1\right)\) such that%

\[
\left \Vert T_{x}-T_{y}\right \Vert ^{2}\leq k\left \Vert x-y\right \Vert
^{2}+\left \Vert \left( I-T\right) x-\left( I-T\right) y\right \Vert
^{2},\forall x,y\in B.
\]

For \(T\) a direct pseudocontractive map , we prove the convergence of Mann iteration to the fixed point of \(T\).

Authors

S.M. Soltuz
Tiberiu Popoviciu Institute of Numerical Analysis

Keywords

pseudocontractive maps; fix ponts; iterative method; convergence

Paper coordinates

Ş.M. Şoltuz, Mann iteration for direct pseudocontractive maps, Bull. Stiint. Univ. Baia Mare, Ser. B, Fasc. Mat.-Inform., 17 (2001) nos. 1-2, 141-144.

PDF

About this paper

Journal

Scientific Bulletin of the University of Baia Mare, Series B, Mathematics-Informatics Fascicola

Publisher Name

University Baia Mare, Romania

DOI
Print ISSN
Online ISSN

3045-1833

google scholar link

??

Paper (preprint) in HTML form

2001-Soltuz-MC-Mann-iteration-b-

MANN ITERATION FOR DIRECT PSEUDOCONTRACTIVE MAPS

Ştefan M. ŞOLTUZ

Abstract

In this note we introduce a new class of maps. Let X X XXX be a real normed space, and B X B X B sub XB \subset XBX be a nonempty set. The map T : B B T : B B T:B rarr BT: B \rightarrow BT:BB is direct pseudocontractive if there exists k ( 0 , 1 ) k ( 0 , 1 ) k in(0,1)k \in(0,1)k(0,1) such that

T x T y 2 k x y 2 + ( I T ) x ( I T ) y 2 , x , y B . T x T y 2 k x y 2 + ( I T ) x ( I T ) y 2 , x , y B . ||Tx-Ty||^(2) <= k||x-y||^(2)+||(I-T)x-(I-T)y||^(2),AA x,y in B.\|T x-T y\|^{2} \leq k\|x-y\|^{2}+\|(I-T) x-(I-T) y\|^{2}, \forall x, y \in B .TxTy2kxy2+(IT)x(IT)y2,x,yB.
For T T TTT a direct pseudocontractive map, we prove the convergence of Mann iteration to the fixed point of T T TTT.
MSC. 47H10
Keywords: preudocontractive maps, fix points, iterative method, convergence
  1. Introduction. Let H H HHH be a real Hilbert space, let B H B H B sub HB \subset HBH be a nonempty, convex set. Let T : B B T : B B T:B rarr BT: B \rightarrow BT:BB be a map. Let x 1 B x 1 B x_(1)in Bx_{1} \in Bx1B, be an arbitrary fixed point. We consider the iteration
(1) x n + 1 = ( 1 α n ) x n + α n T x n (1) x n + 1 = 1 α n x n + α n T x n {:(1)x_(n+1)=(1-alpha_(n))x_(n)+alpha_(n)Tx_(n):}\begin{equation*} x_{n+1}=\left(1-\alpha_{n}\right) x_{n}+\alpha_{n} T x_{n} \tag{1} \end{equation*}(1)xn+1=(1αn)xn+αnTxn
The sequence ( α n ) n 1 α n n 1 (alpha_(n))_(n >= 1)\left(\alpha_{n}\right)_{n \geq 1}(αn)n1 satisfics: ( α n ) n 1 ( 0 , 1 ) , n = 1 α n = α n n 1 ( 0 , 1 ) , n = 1 α n = (alpha_(n))_(n >= 1)sub(0,1),sum_(n=1)^(oo)alpha_(n)=oo\left(\alpha_{n}\right)_{n \geq 1} \subset(0,1), \sum_{n=1}^{\infty} \alpha_{n}=\infty(αn)n1(0,1),n=1αn=, and n = 1 α n 2 < n = 1 α n 2 < sum_(n=1)^(oo)alpha_(n)^(2) < oo\sum_{n=1}^{\infty} \alpha_{n}^{2}< \inftyn=1αn2<. The last relation implies that lim n α n = 0 lim n α n = 0 lim_(n rarr oo)alpha_(n)=0\lim _{n \rightarrow \infty} \alpha_{n}=0limnαn=0. A prototype for ( α n ) n 1 α n n 1 (alpha_(n))_(n >= 1)\left(\alpha_{n}\right)_{n \geq 1}(αn)n1 is ( 1 / n ) n 1 ( 1 / n ) n 1 (1//n)_(n >= 1)(1 / n)_{n \geq 1}(1/n)n1.
Definition 1 The map T T TTT is called pseudocontractive if
T x T y 2 x y 2 + ( I T ) x ( I T ) y 2 , x , y B . T x T y 2 x y 2 + ( I T ) x ( I T ) y 2 , x , y B . ||Tx-Ty||^(2) <= ||x-y||^(2)+||(I-T)x-(I-T)y||^(2),AA x,y in B.\|T x-T y\|^{2} \leq\|x-y\|^{2}+\|(I-T) x-(I-T) y\|^{2}, \forall x, y \in B .TxTy2xy2+(IT)x(IT)y2,x,yB.
In [7] we can see an example of a Lipechitz pseudocontractive map with a unique fixed point for which every non trivial Mann sequence fails to converge. The set B B BBB is nonempty, convex and compact.
Definition 2 The map T T TTT is called strongly pseudocontractive if there exists q ( 0 , 1 ) q ( 0 , 1 ) q in(0,1)q \in(0,1)q(0,1) such that
T x T y 2 x y 2 + q ( I T ) x ( I T ) y 2 , x , y B . T x T y 2 x y 2 + q ( I T ) x ( I T ) y 2 , x , y B . ||Tx-Ty||^(2) <= ||x-y||^(2)+q||(I-T)x-(I-T)y||^(2),AA x,y in B.\|T x-T y\|^{2} \leq\|x-y\|^{2}+q\|(I-T) x-(I-T) y\|^{2}, \forall x, y \in B .TxTy2xy2+q(IT)x(IT)y2,x,yB.
In [1], [2], [3], [5], [8], [11] the map T T TTT is considered strongly pseudocontractive. The sequence ( x n ) n 1 x n n 1 (x_(n))_(n >= 1)\left(x_{n}\right)_{n \geq 1}(xn)n1 given by (1) strongly converges to a fixed point of T T TTT.
We introduce the following class of maps:
Definition 3 The map T T TTT is called direct pseudocontractive if there exists k ( 0 , 1 ) k ( 0 , 1 ) k in(0,1)k \in(0,1)k(0,1) such that
(2) T x T y 2 k x y 2 + ( I T ) x ( I T ) y 2 , x , y B . (2) T x T y 2 k x y 2 + ( I T ) x ( I T ) y 2 , x , y B . {:(2)||Tx-Ty||^(2) <= k||x-y||^(2)+||(I-T)x-(I-T)y||^(2)","AA x","y in B.:}\begin{equation*} \|T x-T y\|^{2} \leq k\|x-y\|^{2}+\|(I-T) x-(I-T) y\|^{2}, \forall x, y \in B . \tag{2} \end{equation*}(2)TxTy2kxy2+(IT)x(IT)y2,x,yB.
The class of direct pseudocontractive maps is nonempty. If T T TTT is a contraction, then T T TTT is a direct pseudocontractive map.Picard -Banach Theorem can't be used to find the fixed point of a direct psendocontractive map. Instead, Mann iteration (1) can be successfully used. Our aim is to give a convergence result for (1). We denote by F ( T ) := { x B : T x = x } F ( T ) := { x B : T x = x } F(T):={x in B:Tx=x}F(T):=\{x \in B: T x=x\}F(T):={xB:Tx=x}.
Remark 1 If T T TTT is a direct pseudocontractive map and has F ( T ) F ( T ) F(T)!=O/F(T) \neq \emptysetF(T), then T T TTT hass a unique fixed point.
Proof. Let x x x^(**)x^{*}x and y y y^(**)y^{*}y be two distinct fixed points. From (2) we have
T x T y 2 k x y 2 x y 2 k x y 2 ( 1 k ) x y 2 0 , k ( 0 , 1 ) T x T y 2 k x y 2 x y 2 k x y 2 ( 1 k ) x y 2 0 , k ( 0 , 1 ) {:[||Tx^(**)-Ty^(**)||^(2) <= k||x^(**)-y^(**)||^(2)],[||x^(**)-y^(**)||^(2) <= k||x^(**)-y^(**)||^(2)],[(1-k)||x^(**)-y^(**)||^(2) <= 0","k in(0","1)]:}\begin{aligned} \left\|T x^{*}-T y^{*}\right\|^{2} & \leq k\left\|x^{*}-y^{*}\right\|^{2} \\ \left\|x^{*}-y^{*}\right\|^{2} & \leq k\left\|x^{*}-y^{*}\right\|^{2} \\ (1-k)\left\|x^{*}-y^{*}\right\|^{2} & \leq 0, k \in(0,1) \end{aligned}TxTy2kxy2xy2kxy2(1k)xy20,k(0,1)
Hence x = y x = y x^(**)=y^(**)x^{*}=y^{*}x=y. Thus F ( T ) = { x } F ( T ) = x F(T)={x^(**)}F(T)=\left\{x^{*}\right\}F(T)={x}.
The following lemma can be found in [10] as Lemma 4. Also, it can be found in [12] as Lemma 1.2, with an other proof. In [1] can be found as Lemma 2, the proof is similar to the proof of Lemma 1 from [8].
Lemma 4 [1], [10], [12] Let ( a n ) n 1 a n n 1 (a_(n))_(n >= 1)\left(a_{n}\right)_{n \geq 1}(an)n1 be a nonnegative sequence which verifies where a n + 1 ( 1 λ n ) a n + σ n , ( λ n ) n 1 ( 0 , 1 ) , n = 1 λ n = a n + 1 1 λ n a n + σ n , λ n n 1 ( 0 , 1 ) , n = 1 λ n = a_(n+1) <= (1-lambda_(n))a_(n)+sigma_(n),(lambda_(n))_(n >= 1)sub(0,1),sum_(n=1)^(oo)lambda_(n)=ooa_{n+1} \leq\left(1-\lambda_{n}\right) a_{n}+\sigma_{n},\left(\lambda_{n}\right)_{n \geq 1} \subset(0,1), \sum_{n=1}^{\infty} \lambda_{n}=\inftyan+1(1λn)an+σn,(λn)n1(0,1),n=1λn= and σ n = o ( λ n ) σ n = o λ n sigma_(n)=o(lambda_(n))\sigma_{n}=o\left(\lambda_{n}\right)σn=o(λn). Then lim n a n = 0 lim n a n = 0 lim_(n-oo)a_(n)=0\lim _{n-\infty} a_{n}=0limnan=0.
The following result is proved in [4].
Lemma 5 (4) Let H H HHH be a Hilbert space, the following relation is true for all x , y H x , y H x,y in Hx, y \in Hx,yH, and for all λ ( 0 , 1 ) λ ( 0 , 1 ) lambda in(0,1)\lambda \in(0,1)λ(0,1) :
(3) ( 1 λ ) x + λ y 2 = ( 1 λ ) x 2 + λ y 2 λ ( 1 λ ) x = y 2 . (3) ( 1 λ ) x + λ y 2 = ( 1 λ ) x 2 + λ y 2 λ ( 1 λ ) x = y 2 . {:(3)||(1-lambda)x+lambda y||^(2)=(1-lambda)||x||^(2)+lambda||y||^(2)-lambda(1-lambda)||x=y||^(2).:}\begin{equation*} \|(1-\lambda) x+\lambda y\|^{2}=(1-\lambda)\|x\|^{2}+\lambda\|y\|^{2}-\lambda(1-\lambda)\|x=y\|^{2} . \tag{3} \end{equation*}(3)(1λ)x+λy2=(1λ)x2+λy2λ(1λ)x=y2.
2.The main result.
We are able now to give the main result:
Theorem 6 Let H H HHH be a real Hilbert space, let B H B H B sub HB \subset HBH be a nonempty, convex, bounded and closed set and let T : B B T : B B T:B rarr BT: B \rightarrow BT:BB be a continuous, direct pseudocontructive map, with F ( T ) F ( T ) F(T)!=O/F(T) \neq \emptysetF(T). Then for each x 1 x 1 x_(1)x_{1}x1 a fixed point in B B BBB, the sequence ( x n ) n 1 x n n 1 (x_(n))_(n >= 1)\left(x_{n}\right)_{n \geq 1}(xn)n1 given by (1) converges strongly to the unique fixed point of T T TTT.
Proof. Let x F ( T ) x F ( T ) x^(**)in F(T)x^{*} \in F(T)xF(T). From remark 2 we know that F ( T ) = { x } F ( T ) = x F(T)={x^(**)}F(T)=\left\{x^{*}\right\}F(T)={x}. Using (2) and (3) we get
x n + 1 x 2 = ( 1 α n ) x n + α n T x n x 2 = ( 1 α n ) ( x n x ) + α n ( T x n x ) 2 = ( 1 α n ) x n x 2 + α n T x n x 2 α n ( 1 α n ) T x n x n 2 ( 1 α n ) x n x 2 + α n k x n x 2 + + α n T x n x n 2 α n ( 1 α n ) T x n x n 2 [ 1 ( 1 k ) α n ] x n x 2 + α n 2 T x n x n 2 . x n + 1 x 2 = 1 α n x n + α n T x n x 2 = 1 α n x n x + α n T x n x 2 = 1 α n x n x 2 + α n T x n x 2 α n 1 α n T x n x n 2 1 α n x n x 2 + α n k x n x 2 + + α n T x n x n 2 α n 1 α n T x n x n 2 1 ( 1 k ) α n x n x 2 + α n 2 T x n x n 2 . {:[||x_(n+1)-x^(**)||^(2)],[=||(1-alpha_(n))x_(n)+alpha_(n)Tx_(n)-x^(**)||^(2)],[=||(1-alpha_(n))(x_(n)-x^(**))+alpha_(n)(Tx_(n)-x^(**))||^(2)],[=(1-alpha_(n))||x_(n)-x^(**)||^(2)+alpha_(n)||Tx_(n)-x^(**)||^(2)-alpha_(n)(1-alpha_(n))||Tx_(n)-x_(n)||^(2)],[ <= (1-alpha_(n))||x_(n)-x^(**)||^(2)+alpha_(n)k||x_(n)-x^(**)||^(2)+],[+alpha_(n)||Tx_(n)-x_(n)||^(2)-alpha_(n)(1-alpha_(n))||Tx_(n)-x_(n)||^(2)],[ <= [1-(1-k)alpha_(n)]||x_(n)-x^(**)||^(2)+alpha_(n)^(2)||Tx_(n)-x_(n)||^(2).]:}\begin{aligned} & \left\|x_{n+1}-x^{*}\right\|^{2} \\ = & \left\|\left(1-\alpha_{n}\right) x_{n}+\alpha_{n} T x_{n}-x^{*}\right\|^{2} \\ = & \left\|\left(1-\alpha_{n}\right)\left(x_{n}-x^{*}\right)+\alpha_{n}\left(T x_{n}-x^{*}\right)\right\|^{2} \\ = & \left(1-\alpha_{n}\right)\left\|x_{n}-x^{*}\right\|^{2}+\alpha_{n}\left\|T x_{n}-x^{*}\right\|^{2}-\alpha_{n}\left(1-\alpha_{n}\right)\left\|T x_{n}-x_{n}\right\|^{2} \\ \leq & \left(1-\alpha_{n}\right)\left\|x_{n}-x^{*}\right\|^{2}+\alpha_{n} k\left\|x_{n}-x^{*}\right\|^{2}+ \\ & +\alpha_{n}\left\|T x_{n}-x_{n}\right\|^{2}-\alpha_{n}\left(1-\alpha_{n}\right)\left\|T x_{n}-x_{n}\right\|^{2} \\ \leq & {\left[1-(1-k) \alpha_{n}\right]\left\|x_{n}-x^{*}\right\|^{2}+\alpha_{n}^{2}\left\|T x_{n}-x_{n}\right\|^{2} . } \end{aligned}xn+1x2=(1αn)xn+αnTxnx2=(1αn)(xnx)+αn(Txnx)2=(1αn)xnx2+αnTxnx2αn(1αn)Txnxn2(1αn)xnx2+αnkxnx2++αnTxnxn2αn(1αn)Txnxn2[1(1k)αn]xnx2+αn2Txnxn2.
The sequence ( T x n x n 2 ) n 1 T x n x n 2 n 1 (||Tx_(n)-x_(n)||^(2))_(n >= 1)\left(\left\|T x_{n}-x_{n}\right\|^{2}\right)_{n \geq 1}(Txnxn2)n1 is bounded, because B B BBB is bounded. There exists M > 0 M > 0 M > 0M>0M>0 such that T x n x n 2 < M T x n x n 2 < M ||Tx_(n)-x_(n)||^(2) < M\left\|T x_{n}-x_{n}\right\|^{2}<MTxnxn2<M, for all n 1 n 1 n >= 1n \geq 1n1. We denote a n := x n x 2 a n := x n x 2 a_(n):=||x_(n)-x^(**)||^(2)a_{n}:= \left\|x_{n}-x^{*}\right\|^{2}an:=xnx2, and we get:
a n + 1 [ 1 ( 1 k ) α n ] a n + α n 2 M . a n + 1 1 ( 1 k ) α n a n + α n 2 M . a_(n+1) <= [1-(1-k)alpha_(n)]a_(n)+alpha_(n)^(2)M.a_{n+1} \leq\left[1-(1-k) \alpha_{n}\right] a_{n}+\alpha_{n}^{2} M .an+1[1(1k)αn]an+αn2M.
Let us denote by
λ n := ( 1 k ) α n σ n := α n 2 M . λ n := ( 1 k ) α n σ n := α n 2 M . {:[lambda_(n):=(1-k)alpha_(n)],[sigma_(n):=alpha_(n)^(2)M.]:}\begin{aligned} & \lambda_{n}:=(1-k) \alpha_{n} \\ & \sigma_{n}:=\alpha_{n}^{2} M . \end{aligned}λn:=(1k)αnσn:=αn2M.
Observe that λ n = ( 1 k ) α n ( 0 , 1 ) λ n = ( 1 k ) α n ( 0 , 1 ) lambda_(n)=(1-k)alpha_(n)sub(0,1)\lambda_{n}=(1-k) \alpha_{n} \subset(0,1)λn=(1k)αn(0,1), for all n 1 n 1 n >= 1n \geq 1n1. We have n 1 λ n = ( 1 k ) n = 1 α n = n 1 λ n = ( 1 k ) n = 1 α n = sum_(n-1)^(oo)lambda_(n)=(1-k)sum_(n=1)^(oo)alpha_(n)=oo\sum_{n-1}^{\infty} \lambda_{n}= (1-k) \sum_{n=1}^{\infty} \alpha_{n}=\inftyn1λn=(1k)n=1αn=. The following relation is true
lim n σ n λ n = lim n α s 2 M ( 1 k ) α n = M 1 k lim n α n = 0 lim n σ n λ n = lim n α s 2 M ( 1 k ) α n = M 1 k lim n α n = 0 lim_(n rarr oo)(sigma_(n))/(lambda_(n))=lim_(n rarr oo)(alpha_(s)^(2)M)/((1-k)alpha_(n))=(M)/(1-k)lim_(n rarr oo)alpha_(n)=0\lim _{n \rightarrow \infty} \frac{\sigma_{n}}{\lambda_{n}}=\lim _{n \rightarrow \infty} \frac{\alpha_{s}^{2} M}{(1-k) \alpha_{n}}=\frac{M}{1-k} \lim _{n \rightarrow \infty} \alpha_{n}=0limnσnλn=limnαs2M(1k)αn=M1klimnαn=0
Thus, we have σ n = o ( λ n ) σ n = o λ n sigma_(n)=o(lambda_(n))\sigma_{n}=o\left(\lambda_{n}\right)σn=o(λn). From Lemma 1 we get lim n a n = 0 lim n a n = 0 lim_(n rarr oo)a_(n)=0\lim _{n \rightarrow \infty} a_{n}=0limnan=0. Hence lim n x n x = 0 lim n x n x = 0 lim_(n rarr oo)||x_(n)-x^(**)||=0\lim _{n \rightarrow \infty}\left\|x_{n}-x^{*}\right\|=0limnxnx=0. The proof is complete.
Using the Schauder fixed point theorem we give the following corollary:
Corollary 7 Let H H HHH be a real Hilbert space, let B H B H B sub HB \subset HBH be a nonempty, convex, compact set and let T : B B T : B B T:B rarr BT: B \rightarrow BT:BB be a continuous, direct pseudocontractive map. Then for exach x 1 x 1 x_(1)x_{1}x1 a fixed point in B B BBB, the sequence ( x n ) n 1 x n n 1 (x_(n))_(n >= 1)\left(x_{n}\right)_{n \geq 1}(xn)n1 given by (1) converges strongly to the unique fixed poinl of T T TTT.

References

[1] S.S.Chang, Y.J. Cho, B.S. Lee, J.S. Jung, S. M. Kang, Iterative Approximations of Fixed Points and Solutions for Strongly Accretive and Strongly Pseudo-contractive Mappings in Banach Spaces, J. Math. Anal. Appl. 224 (1998), 149-165.
[2] C. E. Chidume, Approximation of Fixed Points of Strongly Pseudocontractive Mappings, Proc. Amer. Math. Soc. 120 (1994), 546-551.
[3] C. E. Chidume, C. Moore, Fixed Point Iteration for Strongly Pseudocontractive Maps, Proc. Amer. Math. Soc. 127 (1999), 1163-1170.
[4] S. Ishikawa, Fized Points by a New Iteration Method, Proc. Amer. Math. Soc. 44 (1974), 147-150.
[5] G. G. Johnson, Fixed Points by Mean value iterations, Proc. Amer. Math. Soc. 34 (1972), 193-195.
[6] R. W. Mann, Mean Value Methods in Iteration, Proc. Amer. Math. Soc. 4 (1953), 504-510.
[7] S.A. Mutangadura, C.E. Chidume, An Example of the Mann Iteration Method for Lipschitz Pseudocontractions, internal report ICTP Trieste (2000), http://www.ictp.trieste.it
[8] J. A. Park, Mann-Ileration for Strictly Pseudocontructive Maps, J. Korean Math. Soc. 31 (1994), 333-337.
[9] R. U. Verma, A Ficed Point Theorem Involving Lipschitzian Generalized Pseudo-contractions, Proc. Royal Irish Acad. 97A (1997), 83-86.
[10] X. Weng, Pired Point Iteration for Local Strictly Pseudocontractive Mapping, Proc. Amer. Math. Soc. 113 (1991), 727-731.
[11] II. Y. Zhou, Stable Iteration Procedures for Strong Pseudocontractions and Nonlinear Equations Involving Accretive Operators without Lipschitz Assumption, J. Math. Anal. Appl. 230 (1999), 1-30.
[12] H. Zhou, J. Yuting, Approximation of Fixed Points of Strongly Pseudocontractive Maps without Lipschitz Assumption, Proc. Amer. Math. Soc.
125 (1997), 1705-1709.
Received: 12.03.2001
"T. Popoviciu" Institute of
Numerical Analysis
Gh. Bilascu 37, P.O. Box 68-1,
3400 Cluj-Napoca, Romania.
2001

Related Posts