• No results found

On common approximate fixed points of monotone nonexpansive semigroups in Banach spaces

N/A
N/A
Protected

Academic year: 2020

Share "On common approximate fixed points of monotone nonexpansive semigroups in Banach spaces"

Copied!
11
0
0

Loading.... (view fulltext now)

Full text

(1)

R E S E A R C H

Open Access

On common approximate fixed points of

monotone nonexpansive semigroups in

Banach spaces

Mostafa Bachar

1*

and Mohamed A Khamsi

2,3

*Correspondence:

[email protected]

1Department of Mathematics,

College of Sciences, King Saud University, Riyadh, Saudi Arabia Full list of author information is available at the end of the article

Abstract

In this paper, we investigate the common approximate fixed points of monotone nonexpansive semigroups of nonlinear mappings{T(t)}t≥0,i.e., a family such that T(0)x=x,T(s+t)x=T(s)◦T(t)x, where the domain is a Banach space. In particular we prove that under suitable conditions, the common approximate fixed points are the same as the common approximate fixed points set of two mappings from the family. Then we give an algorithm of how to construct an approximate fixed point sequence of the semigroup in the case of a uniformly convex Banach space.

MSC: Primary 47H09; secondary 46B20; 47H10; 47E10

Keywords: approximate fixed point; common approximate fixed point; fixed point; common fixed point; integral equation; monotone nonexpansive mapping;

semigroup

1 Introduction

Nonexpansive mappings are those maps which have Lipschitz constant equal to . The fixed point theory for such mappings is rich and varied. It finds many applications in non-linear functional analysis. The existence of fixed points for nonexpansive mappings in Ba-nach and metric spaces has been investigated since the early s; see,e.g., Belluce and Kirk [, ], Browder [], Bruck [], Lim [].

In recent years, a new direction has been very active essentially after the publication of Ran and Reurings fixed point theorem [] dealing with the extension of the Banach contraction principle to metric spaces endowed with a partial order. In particular, they show how this extension is useful when dealing with some special matrix equations. It is worth mentioning that similar results were discovered by Turinici [, ]. Another similar approach was carried out by Nieto and Rodríguez-López [] and used such arguments in solving some differential equations. In [] Jachymski gave a more general unified version of these extensions by considering graphs instead of a partial order.

(2)

mappings forms a semigroup ifT()x=xandT(s+t)x=T(s)T(t). Such a situation is quite typical in mathematics and applications. For instance, in the theory of dynamical systems, the vector function space would define the state space and the mapping (t,x)→T(t)x

would represent the evolution function of a dynamical system. In the setting of this pa-per, the state space is a Banach space. Our approach is new and different from the ideas found in [–, ]. Indeed when one deals with a contraction, the focus is usually on the distance being complete. But when dealing with nonexpansive mappings, then geometric properties of the space are crucial.

For more on metric fixed point theory, the reader may consult the books [, ]. As for the semigroup theory, we suggest the references [–].

2 Preliminaries

Let (X,·) be a Banach vector space and suppose thatis a partial order onX. Through-out, we will assume that the partial orderand the liner structure ofXenjoys the follow-ing convexity property:

abandcdαa+ ( –α)cαb+ ( –α)d

for anyα∈[, ] anda,b,c,dX. Moreover, we will assume that order intervals are closed. Recall that an order interval is any of the subsets

(i) [a,→) ={xX;ax}, (ii) (←,a] ={xX;xa},

for anyaX. Next we give the definition of monotone nonexpansive mappings.

Definition . Let (X, · ,) be as above. LetC be a nonempty subset ofX. A map

T:BXis said to be

(a) monotone ifT(x)T(y)wheneverxy; (b) monotone nonexpansive ifTis monotone and

T(x) –T(y)≤ xy

for anyx,yCsuch thatxy.

A fixed point ofT is any elementxCsuch thatT(x) =x. The set of all fixed points of

T is denoted byFix(T). A sequence{xn}is called an approximate fixed point sequence of

T iflimn→∞T(xn) –xn= . The set of approximate fixed point sequences ofT will be denoted byAFPS(T).

This definition is extended to the case of semigroup of mappings.

Definition . Let (X, · ,) be as above. LetC be a nonempty subset ofX. A one-parameter familyF={T(t);t≥}of mappings fromCintoCis said to be a monotone nonexpansive semigroup ifFsatisfies the following conditions:

(i) T()x=xforxC;

(ii) T(t+s) =T(t)◦T(s)fort,s∈[,∞);

(3)

Define then the set of all common fixed points ofFasFix(F) =t≥Fix(T(t)). Similarly, define the set of approximate point sequences ofF, denoted byAFPS(F), asAFPS(F) =

t≥AFPS(T(t)).

Next we give an example of such semigroup.

Example . Let (X, · ,) be as above. LetCbe a nonempty closed, bounded convex subset ofX. LetJ:CCbe a monotone nonexpansive mapping. LetxCbe such that

xJ(x). Consider the recurrent sequence defined by

u(t) =x,

un+(t) =etx+

t

estJ(un(s))ds

(.)

for anyt∈[,A], whereAis a fixed positive number. Then the sequence{un(t)}is Cauchy for anyt∈[,A]. Indeed, let us suppose thatx,y: [,A]→Xare continuous functions. For eacht∈[,A], we have

ety(t) +

t

estx(s)dsety∞+K(t)x∞, (.)

whereu∞=sup{u(s);t∈[,A]}andK(t) =testds=  –et. Indeed, lett[,A] be fixed, andτ ={ti;i= , , . . . ,n}be any subdivision of [,t]. Set

=ety(t) +

n–

i=

(ti+–ti)etitx(ti).

The family{}is norm-convergent to

S=ety(t) + t

estx(s)ds,

when|τ|=sup{|ti+–ti|;i= , , . . . , (n– )}goes to . Since

sup

s∈[,t]

x(s)≤ sup

s∈[,A]

x(s)=x

and

ety(t)+

n–

i=

(ti+–ti)etitx∞≤ety(t)+K(t)x∞,

where we used

n–

i=

(ti+–ti)etitt

estds=K(t),

then we haveSety(t)+K(t)x. Therefore we have

ety(t) + t

(4)

for anyt∈[,A]. Next we discuss the sequence{un(t)}. Since

u(t) =etx+

t

estJ(x)ds=etx+ –et J(x), t∈[,A],

we haveu(t)u(t) for anyt∈[,A]. Assume thatun(t)un+(t) for anyt∈[,A]. Then

we haveJ(un(t))J(un+(t)) for anyt∈[,A] becauseJis monotone. Using the properties

of the partial order, we get

un+(t) =etx+

t

estJun(s) dsetx+ t

estJun+(s) ds=un+(t)

for anyt∈[,A]. By induction, we conclude that{un(t)}is monotone increasing for any

t∈[,A]. Moreover, we have

un+(t) –un+(t) =

t

estJun+(s) –J

un(s)

ds, t∈[,A],

which implies by using the monotone nonexpansive behavior ofJ

un+(t) –un+(t)≤

t

estJun+(s) –J

un(s) ds, t∈[,A], or

un+(t) –un+(t)≤

t

estun+(s) –un(s)ds, t∈[,A]

for anyn≥. From this inequality and by induction, we get un+(t) –un(t)≤

 –eA ndiam(C), t∈[,A],

wherediam(C) =sup{yz;y,zC}<∞, sinceCis bounded. Hence un+h(t) –un(t)≤

( –eA)n

 – ( –eA)diam(C) =e

A –eA n

diam(C)

for anyt∈[,A] andn,h∈N. Clearly this implies that{un(t)}is Cauchy for anyt∈[,A]. Since order intervals are closed, we conclude that the limitu(t) of{un(t)}satisfiesun(t)

u(t) for anyn≥ andt∈[,A]. Using the monotone nonexpansive behavior ofJ, we get Jun(t) –J

u(t) ≤un(t) –u(t)≤eA

 –eA ndiam(C)

for anyt∈[,A] andn,h∈N. In particular, we havelimn→+∞J(un(t)) =J(u(t)), uniformly in [,A], which implies

u(t) =etx+ t

estJu(s) ds, t[,A]. (.)

(5)

F: [,∞)×CCby

T(t)x=u(t).

ThenFdefines a semigroup which is monotone nonexpansive. Indeed, we haveu(t) –

U(t)=xy=xy∞. Assumeun(t) –Un(t) ≤ xy, then by using (.) we will have

un+(t) –Un+(t)≤etxy+

t

estun(s) –Un(s)dsxy.

By induction,un(t) –Un(t) ≤ xyholds for anyn∈Nandt∈[,A]. Hence T(t)xT(t)y=u(t) –U(t)≤ xy

for anyt∈[,A]. Moreover, we have

T(t+s)x=e–(t+s)x+ t+s

e–(t+sσ)JT(σ)x

=e–(t+s)x+

t

e–(t+sσ)JT(σ)x dσ+

t+s

t

e–(t+sσ)JT(σ)x dσ,

=es

etx+ t

e–(tσ)JT(σ)x

+

s

e–(sσ)JT(t+σ)x

=esT(t)x+ s

e–(sσ)JT(t+σ)x dσ

=T(s)T(t)x

for anyt≥. Note that if we started by a pointxCsuch thatJ(x)x, then we would have found that the sequence{un(t)}is monotone decreasing for anyt≥. In other words, the conclusion would have been the same. Moreover, ifxis a fixed point ofJ, then we have

un(t) =xfor anyn∈Nandt≥. HenceT(t)x=xfor anyt≥,i.e.,x

t≥Fix(T(t)). On the other hand, it is easy to show that ifx∈Fix(F), thenx∈Fix(J), which means we have

Fix(J) = t≥

FixT(t) .

3 Common approximate fixed points of semigroups

Before we state our first result, we need the following definition.

Definition . Let (X, · ) be a Banach space andCXbe nonempty. A one-parameter familyF={T(t);t≥}of mappings fromCintoXis said to be:

(i) continuous onCif for anyxC, the mappingtT(t)xis continuous,i.e., for any

t≥, we have lim

tt

T(t)xT(t)x= 

(6)

(ii) strongly continuous onCif for any bounded nonempty subsetKC, we have

lim

tt

sup

xK

T(t)xT(t)x = .

The following technical lemmas will be useful throughout.

Lemma . Let(X, · )be a Banach space and CX be nonempty.Let J:CC be a uniformly continuous mapping.Then we haveAFPS(J)⊂AFPS(Jm)for any m.

Proof Without loss of generality we may assume thatAFPS(J) is not empty. Let{xn} ∈

AFPS(J),i.e.,limn→∞J(xn) –xn= . Fixm≥, then we have Jm(xn) –xn

m–

k=

Jk+(xn) –Jk(xn)

for anyn≥. SinceTis uniformly continuous, then we must have

lim

n→∞J k+(x

n) –Jk(xn)= 

for anyk≥. Sincemis fixed, we getlimn→∞Jm(xn) –xn= ,i.e.,{xn} ∈AFPS(Jm).

Lemma . Let(X, · )be a Banach space and CX be nonempty.LetF={T(t);t≥}

be a one-parameter semigroup of uniformly continuous mappings from C into C.Letαand

βbe two positive real numbers.Then we have

AFPST(α) ∩AFPST(β) ⊂ tG+(α,β)

AFPST(t),

where G+(α,β) ={+≥;m,k∈Z}.

Proof Without loss of generality we may assume thatAFPS(T(α))∩AFPS(T(β)) is not empty. Let{xn} ∈AFPS(T(α))∩AFPS(T(β)). LettG+(α,β). Then we have two cases.

First assumet=+, wherem,k≥. Then

T(t)xn=T(+)xn=Tm(α)

Tk(β)xn ,

which implies

T(t)xnxnTm(α)

Tk(β)xnTm(α)xn+Tm(α)xnxn

for any n ≥ . Using Lemma . and the uniform continuity of Tm(α), we get

limn→∞T(t)xnxn= ,i.e.,{xn} ∈AFPS(T(t)). Next assume thatt=+, where eithermorkis negative. Without loss of generality assumet=, wherem,k≥. We have

(7)

SinceT() =T()T(), andT() is uniformly continuous, Lemma . implieslimn→∞T(t)xnxn= ,i.e.,{xn} ∈AFPS(T(t)). Hence

{xn} ∈

tG+(α,β)

AFPST(t) .

The following lemma, which can be found in any introductory course on real analysis, will be crucial to proving the first result on common approximate fixed point of semi-groups.

Lemma .[] Let G be a nonempty additive subgroup ofR.Then G is either dense in

Ror there exists a> such that G=a·Z={an,n∈Z}.Therefore ifαandβare two real numbers such that α

βis irrational,then the set

G(α,β) ={αn+βm;n,m∈Z}

is dense inR.In particular,the set G+(α,β) =G(α,β)∩[, +∞)is dense in[, +∞). Theorem . Let(X, · )be a Banach space and CX be nonempty and bounded.Let

F={T(t);t≥}be a one-parameter semigroup of uniformly continuous mappings from C into C.Assume thatFis strongly continuous.Letαandβbe two positive real numbers such thatαβ is irrational.Then we have

AFPST(α) ∩AFPST(β) =AFPS(F).

Proof Since AFPS(F) ⊂ AFPS(T(α)) ∩ AFPS(T(β)), it is enough to just prove

AFPS(T(α)) ∩ AFPS(T(β)) ⊂ AFPS(F). Without loss of generality, assume that

AFPS(T(α))∩AFPS(T(β)) is not empty. Let{xn} ∈AFPS(T(α))∩AFPS(T(β)). Lemma . implies that

{xn} ∈

tG+(α,β)

AFPST(t) .

From Lemma ., we know thatG+(α,β) =G(α,β)∩[, +∞) is dense in [, +∞). Lett

[, +∞). Then there existstmG+(α,β),m≥, such thatlimm→∞tm=t. We have xnT(t)xnxnT(tm)xn+T(tm)xnT(t)xn

xnT(tm)xn+sup xC

T(t)xT(t)x.

Letε> . SinceF is strongly continuous, there existsm≥ such that for anymm,

we have

sup

xC

T(tm)xT(t)x<ε.

Since{xn} ∈AFPS(T(m)) from Lemma ., there existsn≥ such that

(8)

for anynn. Hence

xnT(t)xnxnT(tm)(xn)+sup

xC

T(tm)xT(t)x< ε

for anynn. Sinceεwas arbitrarily positive, we conclude thatlimn→∞T(t)xnxn= ,

i.e.,{xn} ∈AFPS(T(t)) for anyt≥. As a corollary to Theorem ., we get the following result.

Corollary . Let(X, · )be a Banach space and CX be nonempty and bounded.Let

F={T(t);t≥}be a one-parameter semigroup of uniformly continuous mappings from C into C.Assume thatFis strongly continuous.Then we have

AFPS(F) =AFPST() ∩AFPST(π) =AFPST() ∩AFPST(√) .

All of the results obtained in this section may be easily stated in metric spaces. In the next section, we give an algorithm of how to construct an approximate fixed point sequence of two maps.

4 Common approximate fixed points of two monotone mappings

In this section we discuss a construction of a common approximate fixed point sequence of two monotone nonexpansive mappings defined on a Banach space (X, · ) endowed with a partial orderas described before. LetCXbe a nonempty convex subset. LetJ,H:

CCbe two mappings. Fixx∈C, Das and Debata [] studied the strong convergence

of Ishikawa iterates{xn}defined by

xn+=αnH

βnJ(xn) + ( –βn)xn + ( –αn)xn, (DD)

whereαn,βn∈[, ]. Under suitable assumptions, we will show that{xn}is an approximate fixed point sequence of bothJandH. Assume thatJandHare monotone nonexpansive. Assume that there existsx∈Csuch thatxJ(x) andxH(x). We will also assume

thatHandJhave a common fixed pointpCsuch thatxandpare comparable. Using

the convexity properties of the partial order, we will easily show thatxnandpare com-parable. SinceH andJare monotone nonexpansive, we getH(xn) –pxnpand

J(xn) –pxnpfor anyn≥. Hence

xn+–p=αnH(yn) + ( –αn)xnp

αnH(yn) –p+ ( –αn)xnp

αnynp+ ( –αn)xnp

=αnβnJ(xn) + ( –βn)xnp+ ( –αn)xnp

αn

βnJ(xn) –p+ ( –βn)xnp

+ ( –αn)xnp

αn

βnxnp+ ( –βn)xnp

+ ( –αn)xnp

(9)

whereyn=βnJ(xn) + ( –βn)xnfor anyn≥. This proves that{xnp}is decreasing, which implies thatlimn→∞xnpexists. Using the above inequalities, we get

lim

n→∞xnp=nlim→∞αnH(yn) + ( –αn)xnp = lim

n→∞

αnH(yn) –p+ ( –αn)xnp

= lim

n→∞

αnynp+ ( –αn)xnp

= lim

n→∞

αnβnJ(xn) + ( –βn)xnp+ ( –αn)xnp

= lim

n→∞

αn

βnJ(xn) –p+ ( –βn)xnp + ( –αn)xnp

.

Before we state the main theorem of this section, let us recall the definition of a uniformly convex Banach space.

Definition .[] Let (X, · ) be a Banach space.Xis said to be uniformly convex if and only if for anyε> , we haveδX(ε) > , where

δX(ε) =inf

 –

x+y;x ≤,y ≤ andxyε

.

It is well known that ifXis uniformly convex, thenXis reflexive []. Moreover, we have []

αx+ ( –α)y≤ –δX

min{α,  –α}ε

for anyα∈(, ),ε>  andx,yXsuch thatx ≤,y ≤ andxyε. The following lemma will be needed to prove the main result of this section.

Lemma .[] Let(X,·)be a uniformly convex Banach space.Let{xn}and{yn}be in X

such thatlim supn→∞xnR,lim supn→∞ynR,andlimn→∞αnxn+ ( –αn)yn=R,

whereαn∈[a,b],with <ab< ,and R≥.Then we have

lim

n→∞xnyn= .

Next we state the main result of this section.

Theorem . Let C be a nonempty,closed and convex subset of a uniformly convex Ba-nach space(X, · ).Let H,J:CC be monotone nonexpansive mappings such that H is uniformly continuous.Assume that there exists x∈C such that xJ(x)and xH(x).

We will also assume that H and J have a common fixed point pC such that xand p

are comparable.Consider the sequence{xn}defined by xand the recurrent formula(DD).

Assume thatαn,βn∈[α,β],with <αβ< ,then

lim

n→∞xnH(xn)=  and nlim→∞xnJ(xn)= ,

(10)

Proof We have already seen that{xnp}is decreasing. Setc=limn→∞xnp. Ifc= , then all the conclusions are trivial. Therefore we will assume thatc> . We have already seen that

lim

n→∞xnp=nlim→∞αnH(yn) + ( –αn)xnp = lim

n→∞

αnH(yn) –p+ ( –αn)xnp

.

We claim thatlimn→∞H(yn) –p=c. Indeed, letUbe a nontrivial ultrafilter overN. Then we havelimn,Uαn=α∞∈[α,β] andlimn,Uxnp=c. Hence

c=αlim

n,U

H(yn) –p+ ( –α∞)c.

Sinceα∞= , we getlimn,UH(yn) –p=c. SinceUwas arbitrary, we getlimn→∞H(yn) –

p=cas claimed. Therefore we have

lim

n→∞xnp=nlim→∞H(yn) –p=nlim→∞αnH(yn) + ( –αn)xnp. Using Lemma ., we getlimn→∞H(yn) –xn= . Since we already proved

lim

n→∞xnp=nlim→∞

αnynp+ ( –αn)xnp

,

a similar argument will show thatlimn→∞ynp=limn→∞xnp. Moreover, we have

lim

n→∞xnp=nlim→∞

αn

βnJ(xn) –p+ ( –βn)xnp + ( –αn)xnp

.

If we use again ultrafilters, one will easily prove thatlimn→∞J(xn) –p=limn→∞xnp. Hence we have

lim

n→∞xnp=nlim→∞J(xn) –p=nlim→∞βnJ(xn) + ( –βn)xnp. Using again Lemma ., we getlimn→∞J(xn) –xn= . Since

lim

n→∞ynxn=nlim→∞βnJ(xn) –xn= ,

and H is uniformly continuous, we get limn→∞H(xn) –H(yn)= . Combined with

limn→∞H(yn) –xn= , we get

lim

n→∞H(xn) –xn= .

Remark . If we assume thatαn=α andβn=β withα,β∈(, ), then under the as-sumptions of Theorem ., we can prove that

xnxn+ and ynyn+

(11)

convexity properties of the partial orderinX, we will conclude that in fact{xn}and{yn} are weakly convergent. It is not clear to us that the weak-limit of{xn}is a fixed point ofH andJ. This will be the case if we have a demi-closed property for monotone nonexpansive mappings. Otherwise, we may assume thatXsatisfies the Opial condition []. In this case the weak-limit of{xn}will be a fixed point ofTandS. As an example of a Banach space which satisfies all of the above assumptions, one may takeX= p,  <p< +.

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

All authors contributed equally to the writing of this paper. All authors read and approved the final manuscript.

Author details

1Department of Mathematics, College of Sciences, King Saud University, Riyadh, Saudi Arabia.2Department of

Mathematical Sciences, The University of Texas at El Paso, El Paso, TX 79968, USA. 3Department of Mathematics and

Statistics, King Fahd University of Petroleum & Minerals, Dhahran, 31261, Saudi Arabia.

Acknowledgements

The authors would like to extend their sincere appreciation to the Deanship of Scientific Research at King Saud University for funding this Research group No. RG-1435-079.

Received: 4 June 2015 Accepted: 13 August 2015 References

1. Belluce, LP, Kirk, WA: Fixed-point theorems for families of contraction mappings. Pac. J. Math.18, 213-217 (1966) 2. Belluce, LP, Kirk, WA: Nonexpansive mappings and fixed-points in Banach spaces. Ill. J. Math.11, 474-479 (1967) 3. Browder, FE: Nonexpansive nonlinear operators in a Banach space. Proc. Natl. Acad. Sci. USA54, 1041-1044 (1965) 4. Bruck, RE: Properties of fixed point sets of non-expansive mappings in Banach spaces. Trans. Am. Math. Soc.179,

251-262 (1973)

5. Lim, TC: A fixed point theorem for families of nonexpansive mappings. Pac. J. Math.53, 487-493 (1974)

6. Ran, ACM, Reurings, MCB: A fixed point theorem in partially ordered sets and some applications to matrix equations. Proc. Am. Math. Soc.132, 1435-1443 (2004)

7. Turinici, M: Abstract comparison principles and multivariable Gronwall-Bellman inequalities. J. Math. Anal. Appl.117, 100-127 (1986)

8. Turinici, M: Fixed points for monotone iteratively local contractions. Demonstr. Math.19, 171-180 (1986) 9. Nieto, JJ, Rodríguez-López, R: Contractive mapping theorems in partially ordered sets and applications to ordinary

differential equations. Order22, 223-239 (2005)

10. Jachymski, J: The contraction principle for mappings on a metric space with a graph. Proc. Am. Math. Soc.136, 1359-1373 (2007)

11. Turinici, M: Ran and Reurings theorems in ordered metric spaces. J. Indian Math. Soc.78, 207-2014 (2011) 12. Goebel, K, Kirk, WA: Topics in Metric Fixed Point Theory. Cambridge Stud. Adv. Math., vol. 28. Cambridge University

Press, Cambridge (1990)

13. Khamsi, MA, Kirk, WA: An Introduction to Metric Spaces and Fixed Point Theory. Wiley, New York (2001) 14. Pavel, NH: Nonlinear Evolution Operators and Semigroups: Applications to Partial Differential Equations. Lecture

Notes in Mathematics. Springer, Berlin (1987)

15. Pazy, A: Semigroups of Linear Operators and Applications to Partial Differential Equations. Applied Mathematical Sciences, vol. 44. Springer, New York (1983)

16. Reich, S, Shoikhet, D: Nonlinear Semigroups, Fixed Points, and Geometry of Domains in Banach Spaces. Imperial College Press, London (2005)

17. Stewart, I, Tall, D: Algebraic Number Theory and Fermat’s Last Theorem, 3rd edn. AK Peters, Wellesley (2001) 18. Das, G, Debata, P: Fixed points of quasi-nonexpansive mappings. Indian J. Pure Appl. Math.17, 1263-1269 (1986) 19. Beauzamy, B: Introduction to Banach Spaces and Their Geometry. Notas de Mathematica, vol. 68. North-Holland,

Amsterdam (1982)

20. Fukhar-ud-din, H, Khamsi, MA: Approximating common fixed points in hyperbolic spaces. Fixed Point Theory Appl. 2014, 113 (2014). doi:10.1186/1687-1812-2014-113

References

Related documents

(2013) and Raftery, Lalic, and Gerland (2014) presented a Bayesian hi- erarchical model (BHM) for projecting male and female life expectancy probabilistically and jointly for

Then the relay switches the circuit to vacuum pump and vacuum pump flushes out air to create vacuum and then the suction cup sticks on the windshield and get locked .For

The inadequate performance of current automated al- gorithms is likely a result of several complementary fac- tors [ 1 , 3 , 7 , 27 , 28 ]: (i) the intrinsic limitations of

Microvascular pericytes contract during cerebral and coronary ischemia and do not relax after re-opening of the occluded artery, causing incomplete reperfusion. However, the

Model controlled for age, sex, socioeconomic status, distance between home and the practice, appointment delay, mean appointment time per patient, number of appointments per

This article describes GBdI’s main research activities, which include applying Fractal Theory concepts to improve analysis and search operations over very large datasets,

Such features were extracted based on the criteria used by Wikipedia Guidelines [Wikipedia 2008] to manually assess the quality of an article.. The features we were divided into