International Journal of Applied Science and Mathematical Theory E- ISSN 2489-009X P-ISSN
2695-1908, Vol. 8 No. 2 2022 www.iiardjournals.org
Approximation of Fixed Points of Contactive
Operators by a Stochastic Iterative Algorithm
Onwukwe Ijioma1, B.C.Ugoala2, V.N.Madukpe3
1,2,3
Department of Mathematics,
Abia State University, Uturu, Nigeria.
D.O.I: 10.56201/ijasmt.v8.no2.2022.pg63.68
ABSTARCT:
We prove a strong convergence result using stochastic iterative scheme with
contractive operators associated with error terms. The iterative process is faster,
simpler which improves other algorithms, and for the class of mapping in a more
global sense.
KEYWORDS: Contractive operator, stochastic algorithm, fixed point, strong
convergence.
1. INTRODUCTION:
Let be a real Hilbert space with the usual inner product 〈 〉 which
induces the norm‖ ‖. Let { } be a non empty, closed and convex subset of
. Let be a non linear mapping of , the set of fixed points of is
defined as { } (1)
In the sense of Brower-Petryshyn, is -strict pseudo contractive if
such that ‖ ‖ ‖ ‖ ‖
‖ (2)
2 PRELIMINARIES:
Our purpose in this work is to prove a convergence result for approximating
fixed points of contractive operator by stochastic approximation, using
iterative process defined in (2). Our approach is different from that of
Mann(see[2]). In this way, our results improve and generalize corresponding
results of (4 and 5).
The problem of finding fixed point of non expansive mappings by Manns
algorithm (see[1]) has been investigated in copious literatures (see[2]).
The Mann iterative is defined by sequence:
(3)
IIARD – International Institute of Academic Research and Development Page 63
International Journal of Applied Science and Mathematical Theory E- ISSN 2489-009X P-ISSN
2695-1908, Vol. 8 No. 2 2022 www.iiardjournals.org
Note that if is a nonexpansive mapping with a fixed point in a
closed and convex subset of a uniformly convex, with a Frechét differentiable
norm , then the sequence generated by Mann’s algorithm converges
weakly to a fixed point of globally (see[2]).
On the other hand, iterative algorithms for - strict pseudo-contractions are
still developing compare to that of nonexpansive mappings. Strict pseudo-
contractions have many applications due to its direct links with inverse
strongly monotone operators. If is a strongly monotone operator, then
is a strict pseudo-contraction as such the problem can be reduced to
locating zeros for in a fixed point problem for (see[3] and [4]). Safer
Hussain Khan in [5] and [6] proved a convergence result for approximating
fixed point of the class of contractive-likeoperators, which improve the
generalize result.
Definition1: A mapping is said to weakly contractive of
on a closed convex subset if there exists a
continuous and increasing function such that
‖ ‖ ‖ ‖ ‖ ‖
Definition 2: A mapping is said to be non expansive on the closed
convex subset if for all ‖ ‖ ‖ ‖. The
class of weakly contractive maps is contained in the class of nonexpansive
maps.
In this work, we formulate approximation of fixed points of contractive
operator by a stochastic iterative process. In Imoru and Olatinwo (see[6]),
gave a general definition: An operator is called contractive-like
operator if there exists a constant [ and a strictly increasing and
continuous function [ [ , with such that for
each ‖ ‖ ‖ ‖ ‖ ‖ . (4)
We introduced a stochastic iteration process by setting
‖ ‖ ‖ ‖
such that ‖ ‖ ‖ ‖ ‖ ‖ , which guarantee
the existence of unique fixed point. With some assumptions, we have
iteration process as ‖ ‖ ‖ ‖ ‖ ‖
∑ (5)
We investigate the stochastic approximation of the form :
(6)
the factor is an infinite-dimensional vector of random
IIARD – International Institute of Academic Research and Development Page 64
International Journal of Applied Science and Mathematical Theory E- ISSN 2489-009X P-ISSN
2695-1908, Vol. 8 No. 2 2022 www.iiardjournals.org
observations of the clearance operator at random points on
the same probability space ( .
3. FORMULATION OF STOCHASTIC APPLICATION ALGORITHM
is a sequence of
independent random vectors with the conditions:
‖ ‖ (see[11])
Let be the expectation operator. If ‖ ‖ , then
〈 〉 〈 〉 Given any linear continuous monotone map
, there exists a continuous, convex and everywhere Frechet
differentiable scalar function
, such that 〈 〉 〈 〉 (7)
The gradient mapping , uniquely determined for a given
coincides with . Finding such that
〈 〉 [7], is equivalent to finding the
unique zero such that (8)
Given that is a monotone operator defined on Hilbert space into itself then
the iterative scheme (9)
converges strongly to a zero of for a suitable { }.
We consider a stochastic approximation algorithm. The usual form of
stochastic approximation algorithm is a minimum point , of a function
is of the form (10)
Where ∫ is the usual probability law of
is approximation of the gradient , and { } is a
sequence of positive scalars that decreases to zero (see [9], [10] and [12]).
The important part of (10) is the gradient approximation . We applied
objective function measurements and requires the minimization of a
continuous and convex function. Given a sequence of random vectors ,
computed from different data points: , for each
such that approximates strongly and in mean square in the sense
that the estimated vectors are minimized as follows:
Definition 3: The sequence of random vectors , is strongly
approximates if ‖ ‖ (11)
and consistent with in mean square if:
‖ ‖ (12)
where the data points.
IIARD – International Institute of Academic Research and Development Page 65
International Journal of Applied Science and Mathematical Theory E- ISSN 2489-009X P-ISSN
2695-1908, Vol. 8 No. 2 2022 www.iiardjournals.org
4. CONVERGENCE OF THE STOCHASTIC APPROXIMATION
ALGORITHM Theorem 1: Let the sequence { } of positive
numbers satisfies the conditions:
(a)
(b) ∑
(c) ∑
Given that { } is a sequence of random vectors satisfying (9) and (10), and
then the stochastic sequence { } in defined iteratively from
by converges strongly to a unique solution {
} almost surely.
Proof: Let ‖ ‖ , then { } is a
sequence of independent random variables. From (9),
thus the sequence of partial sums ∑ is a
Martingale. But ∑ ∑ ‖ ‖
∑ . The convergence of the series ∑ in (Theorem
1(c)),implies∑ . Thus by Martingale convergence
theorem (see [9]), we have ∑ , so that ‖
‖ .
It is obvious that the stochastic analog of (7) under suitable hypotheses,
behavesAsymptotically as (7), and it implies the convergence properties of
(7) are preservedwhen is replaced by . It follows that the stochastic
iteration scheme: converges strongly to unique solution
{ } From (10) strongly approximates .
Furthermore, it can be established that ‖ ‖ ‖ ‖ for
all least-square approximations. Then the stochastic iterative scheme
generated by the stochastic sequence { } is then given as follows:
Let be an estimate of
(a) Compute
(b) Compute
(c) Compute
(d) Check for convergence:
Is ‖ ‖
Yes:
No: Set and return to (a)
IIARD – International Institute of Academic Research and Development Page 66
International Journal of Applied Science and Mathematical Theory E- ISSN 2489-009X P-ISSN
2695-1908, Vol. 8 No. 2 2022 www.iiardjournals.org
5. Conclusion: Our aim in this paper is to prove a convergence result for
approximating fixed points of a class of contractive-like operators defined in
(4) using the iterative process (7). We prove a strong convergence result
using stochastic iterative scheme with contractive operators associated with
error terms. The iterative process is faster and simpler which improves
other algorithms, and for the class of mapping in a more global sense.
6.Competing Interests: Author(s) have declared that no competing interests exist.
References:
[1] Mann, WR “ Mean Value Methods in Iteration. Proc Am. Math. Soc.
4, 5 06-510 (1953)
[2] Reich, S: Weak Convergence Theorem for non expansive mappings
in Banach spaces. J. Math. Anal. Appl. 67, 274-276 (1979)
[3] Shahazad, N. Zegeye, H : Approximating of common point of fixed
points of a pseudo-contractive mapping and zero of sum of monotone
mappings.Fixed point Theory Appl.,85.(2014)
[4] Zamfirescu, T.: Fixed point Theorems in metric spaces, Arch.Math
(Basel, 23(1972), 292-298
[5] Khan, Safeer. H. Fixed Points of Contractive-Like Operators by a
Faster Iterative Process, Int. Journal of Math. Compl, Nat.
and Phy Eng. Vol 7, No.: 12, 2013
[6] C.O. Imoru, M.O.Olatinwo, On the stability of Picard and Mann
Iteration processes, Carpathian Journal of Mathematics ,
19,155-160, 2003
[7] Browder, F.E., Non linear mapping of nonexpansive and accretive
type in Banach spaces. Bull. Am. Math. Soc. 73,875-882 (1967).
[8] Chidume, C.E. 1987. Iterative approximation of fixed point of
Lipschitzian strictly Pseudo contractive mappings. Proc. Am. Math.
Soc., 99: 283-288.
[9] O. Ijioma and A.C. Okoroafor “A Stochastic Approximation Method
Applied To A WalrasianEquilibrium Model” IJAM vol. 5,No. 1, pp
35-44, 2012.
IIARD – International Institute of Academic Research and Development Page 67
International Journal of Applied Science and Mathematical Theory E- ISSN 2489-009X P-ISSN
2695-1908, Vol. 8 No. 2 2022 www.iiardjournals.org
[10] O. Ijioma“ A Stochastic Iteration Method for A Class of Monotone
Variational Inequalities In Hilbert Space” International Journal of
Scientific and Statistical Computing (IJSSC), vol.(8), Issue(1): 2020
[11] L. Ljung, G. Pfugand H. Walk. “ Stochastic Approximation and
Optimization of Random Systems”
[12] Ya.I.Alber, C.E. Chidume, Jinlu Li. “Stochastic Approximation
Method for Fixed Point Problems” Applied Mathematics,
2012,3,2123-2132 (http://dx.doi.org/10.4236/am.212.312A293 )
IIARD – International Institute of Academic Research and Development Page 68
