R E S E A R C H
Open Access
Mean ergodic theorem for semigroups of
linear operators in multi-Banach spaces
Hassan Morghi Kenari
1, Reza Saadati
2, Mahdi Azhini
1and Yeol Je Cho
3**Correspondence: [email protected] 3Department of Mathematics
Education and the RINS, Gyeongsang National University, Jinju, 660-701, Korea
Full list of author information is available at the end of the article
Abstract
In this paper, by using Rodé’s method, we extend Yosida’s theorem to semigroups of linear operators in multi-Banach spaces.
MSC: Primary 39A10; 39B72; secondary 47H10; 46B03
Keywords: ergodic theorem; semigroups; multi-Banach space
1 Introduction
In , Yosida [] proved the following mean ergodic theorem for linear operators: LetE be a real Banach space andTj(j= , , . . .) be linear operators ofEinto itself such that there
exists a constantCwith(Tn, . . . ,Tjn) ≤Cforn= , , , . . . , andTjis weakly completely
continuous,i.e.,Tjmaps the closed unite ball ofEinto a weakly compact subset ofE. Then
the Cesaro means
Sn,jx=
n
n
k=
Tjkx
converges strongly asn→+∞to a fixed point ofTjfor eachx∈E.
On the other hand, in , Baillon [] proved the following nonlinear ergodic theorem. LetXbe a Banach space andCbe a closed convex subset ofX. The mappingsTj:C→C
(j= , , . . .) are called nonexpansive onCif
Tjx–Tjy ≤ x–y ∀x,y∈C.
LetF(Tj) be the set of fixed points ofTj. IfXis strictly convex,F(Tj) is closed and convex.
In [], Baillon proved the first nonlinear ergodic theorem such that ifXis a real Hilbert space andF(Tj)=∅, then for eachx∈C, the sequence{Sn,jx}defined by
Sn,jx=
n
x+Tjx+· · ·+Tjn–x
converges weakly to a fixed point ofTj. It was also shown by Pazy [] that ifXis a real
Hilbert space andSn,jxconverges weakly toy∈C, theny∈F(Tj).
Recently, Rodé [] and Takahashi [] tried to extend this nonlinear ergodic theorem to semigroup, generalizing the Cesaro means onN={, , . . .}, such that the corresponding
sequence of mappings converges to a projection onto the set of common fixed points. In this paper, by using Rodé’s method, we extend Yosida’s theorem to semigroups of linear operators in multi-Banach spaces. The proofs employ the methods of Yosida [], Rodé [], Greenleaf [] and Takahashi [, ]. Our paper is motivated from ideas in [].
2 Multi-Banach spaces
The notion of multi-normed space was introduced by Dales and Polyakov in []. This concept is somewhat similar to operator sequence space and has some connections with operator spaces and Banach lattices. Motivations for the study of multi-normed spaces and many examples are given in [–].
Let (E, · ) be a complex normed space, and letk∈N. We denote byEkthe linear space
E⊕ · · · ⊕Econsisting ofk-tuples (x, . . . ,xk), wherex, . . . ,xk∈E. The linear operations
onEkare defined coordinate-wise. The zero element of eitherEorEkis denoted by . We denote byNkthe set{, , . . . ,k}and bykthe group of permutations onksymbols.
Definition . LetEbe a linear space, and takek∈N. Forσ∈k, define
Aσ(x) = (xσ(), . . . ,xσ(k)), x= (x, . . . ,xk)∈Ek.
Forα= (αi)∈Ck, define
Mα(x) = (αixi), x= (x, . . . ,xk)∈Ek.
Definition . Let (E, · ) be complex (respectively, real) normed space, and taken∈N. A multi-norm of levelnon{Ek:k∈N
n}is a sequence ( · k:k∈Nn) such that · is a
norm onEkfor eachk∈Nn, such thatx=xfor eachx∈E(so that · is the initial
norm), and such that the following axioms (A)-(A) are satisfied for eachk∈Nnwith
k≥:
(A) for eachσ∈kandx∈Ek, we have Aσ(x)
k=xk;
(A) for eachα, . . . ,αk∈C(respectively, eachα, . . . ,αk∈R) andx∈Ek, we have Mα(x)k≤
max
i∈Nk
|αi| xk;
(A) for eachx, . . . ,xk–, we have
(x, . . . ,xk–, )k=(x, . . . ,xk–)k–;
(A) for eachx, . . . ,xk–∈E
(x, . . . ,xk–,xk–,xk–)k=(x, . . . ,xk–,xk–)k–.
In this case, ((Ek, ·
A multi-norm on{Ek:k∈N}is a sequence
· k
= · k:k∈N
such that ( · k:k∈Nn) is a multi-norm of levelnfor eachn∈N. In this case, ((En, · n) :
n∈N) is a multi-normed space.
Lemma .[] Suppose that((Ek, ·
k) :k∈N)is a multi-normed space,and take k∈
Nn.Then
(a) (x, . . . ,x)k=x(x∈E);
(b) maxi∈Nkxi ≤ (x, . . . ,xk)k≤ k
i=xi ≤kmaxi∈Nkxi(x, . . . ,xk∈E).
It follows from (b) that, if (E, · ) is a Banach space, then (Ek, ·
k) is a Banach space
for eachk∈N; in this case ((Ek, · k) :k∈N) is a multi-Banach space.
Now we state two important examples of multi-norms for an arbitrary normed spaceE; cf.[].
Example . The sequence ( · k:k∈N) on{Ek:k∈N}defined by (x, . . . ,xk)k:=max
i∈Nk
xi (x, . . . ,xk∈E)
is a multi-norm called the minimum multi-norm. The terminology ‘minimum’ is justified by property (b).
Example . Let{( · α
k:k∈N) :α∈A}be the (non-empty) family of all multi-norms on
{Ek:k∈N}. Fork∈N, set (x, . . . ,xk)k:=sup
α∈A
(x, . . . ,xk) α
k (x, . . . ,xk∈E).
Then ( · k:k∈N) is a multi-norm on{Ek:k∈N}, called the maximum multi-norm.
We need the following observation, which can easily be deduced from the triangle in-equality for the norm · kand the property (b) of multi-norms.
Lemma . Suppose that k∈Nand(x, . . . ,xk)∈Ek.For each j∈ {, . . . ,k},let(xjn)n=,,...
be a sequence in E such thatlimn→∞xjn=xj.Then for each(y, . . . ,yk)∈Ekwe have
lim
n→∞
xn–y, . . . ,xkn–yk
= (x–y, . . . ,xk–yk).
Definition . Let ((Ek, ·
k) :k∈N) be a multi-normed space. A sequence (xn) inEis
amulti-nullsequence if, for eachε> , there existsn∈Nsuch that
sup
k∈N
(xn, . . . ,xn+k–)k<ε (n≥n).
Letx∈E. We say that the sequence (xn) ismulti-convergenttox∈Eand write
lim
n→∞xn=x
3 Preliminaries and lemmas
LetEa real Banach space and letE∗be the conjugate space ofE, that is, the space of all continuous linear functionals onE. The value ofx∗∈E∗atx∈Ewill be denoted byx,x∗. We denote bycoDthe convex hull ofD,coDthe closure ofcoD.
LetUbe a linear continuous operator ofEinto itself. Then we denote byU∗the conju-gate operator ofU.
Assumption (A) Let (Ej, ·
j) be a multi-Banach space and{Tj,t:t∈G}(j= , , . . .) be a
family of linear continuous operators of a real Banach spaceEjinto itself such that there
exists a real number Cwith (T,t, . . . ,Tj,t)j≤C for allt∈Gand the weak closure of
{Tj,tx:t∈G}is weakly compact, for eachx∈E. The index setGis a topological semigroup
such thatTj,st=Tj,s·Tj,t for alls,t∈GandTj is continuous with respect to the weak
operator topology:Tj,sx,x∗ → Tj,tx,x∗for allx∈Eandx∗∈E∗ifs→tinG.
We denote bymj(G) the Banach space of all bounded continuous real valued functions
on the topological semigroupGwith the supremum norm. For eachs∈Gandfj∈mj(G),
we define elementslsfjandrsfjinmj(G) given bylsfj(t) =fj(st) andrsfj(t) =fj(ts) for allt∈G.
An elementμj∈mj(G)∗ (the conjugate space ofmj(G)) is called a mean onGifμj=
μj() = moreover, we have(μ, . . . ,μj)j=
j i=μi()
j = . A meanμjonGis called left
(right) invariant ifμj(lsfj) =μj(fj) (μj(rsfj) =μj(fj)) for allfj∈mj(G) ands∈G. An invariant
mean is a left and right invariant mean. We know thatμj∈mj(G)∗is a mean onGif and
only if
inffj(t) :t∈G
≤μj(fj)≤sup
fj(t) :t∈G
for everyfj∈mj(G); see [, –].
Let {Tj,t :t∈G} be a family of linear continuous operators ofE into itself satisfying
Assumption (A) and μj be a mean on G. Fix x∈E. Then, forx∗∈E∗, the real valued
functiont→ Tj,tx,x∗is inmj(G). Denote byμj,tTj,tx,x∗the value ofμjat this function.
By linearity ofμjand of·,·, this is linear inx∗; moreover, since μ,t
T,tx,x∗
, . . . ,μj,t
Tj,tx,x∗
≤(μ, . . . ,μj)j·sup t
T,tx,x∗
, . . . ,μj,t
Tj,tx,x∗
≤sup
t
(Tx, . . . ,Tjx)j·x∗j
≤C· xj·x∗j
it is continuous inx∗. Hence we find thatμj,tTj,tx,·is an element ofE∗∗. So, from weak
compactness ofco{Tj,tx:t∈G}such thatμj,tTj,tx,x∗=Tj,μjx,x
∗for everyx∗∈E∗.
PutK=co{Tj,tx:t∈G}and suppose that the elementμj,tTj,tx,·is not contained in the
n(K), wherenis the natural embedding of the Banach spaceEinto its second conjugate spaceE∗∗. Then, since the convex setn(K) is compact in theweak∗topology ofE∗∗, there exists an elementy∗∈E∗such that
μj,t
Tj,tx,y∗
Hence, we have
μj,t
Tj,tx,y∗
<infy∗,z∗∗:z∗∗∈n(k)
≤infTj,tx,y∗
:t∈G
≤μj,t
Tj,tx,y∗
.
This is a contradiction. Thus, for a meanμjonG, we can define a linear continuous
op-eratorTj,μj ofEinto itself such that(T,μ, . . . ,Tj,μj)j≤C,Tj,μjx∈co{Tj,tx:t∈G}for all
x∈E, andμj,tTj,tx,x∗=Tj,μjx,x∗for allx∈Eandx∗∈E∗. We denote byFj(G) the set
all common fixed points of the mappingsTj,t,t∈G.
Lemma . Assume that a left invariant meanμjexists on G,then Tj,μj(E)⊂Fj(G).
Espe-cially,Fj(G)is then not empty.
Proof Letx∈Eandμbe a left invariant mean onG. Then since, fors∈Gandx∗,
Tj,sTj,μJx,x
∗=T
j,μjx,T
∗
j,sx∗
=μj,t
Tj,tx,Tj,s∗x∗
=μj,t
Tj,sTj,tx,x∗
=μj,t
Tj,stx,x∗
=μj,t
Tj,tx,x∗
=Tj,μjx,x
∗,
we haveTj,sTj,μjx=Tj,μjx. Hence,Tj,μj(E)⊂Fj(G).
Lemma . Letλjbe an invariant mean on G.Then Tj,λjTj,s=Tj,sTj,λj=Tj,λjfor each s∈G
and Tj,λjTj,μj=Tj,μjTj,λj =Tj,λj for each meanμjon G.Especially,Tj,λj is a projection of E
onto F(G).
Proof Lets∈G. Then, since
Tj,λjTj,sx,x
∗=λ
j,t
Tj,tTj,sx,x∗
=λj,t
Tj,tsx,x∗
=λj,t
Tj,tx,x∗
=Tj,λjx,x
∗
forx∈Eandx∗∈E∗, we haveTj,λjTj,s=Tj,λj. It is obvious from Lemma . thatTj,sTj,λj=
Tj,λj for eachs∈G. Letμjbe a mean onG. Then, since
Tj,μjTj,λjx,x∗
=μj,t
Tj,tTj,λjx,x∗
=μj,tTj,λjx,x∗
=Tj,λjx,x
∗
and
Tj,λjTj,μjx,x
∗=T
j,μjx,T
∗
j,λjx
∗=μ
j,t
Tj,tx,Tj,∗λjx
∗
=μj,t
Tj,λjTj,tx,x
∗=μ
j,t
Tj,λjx,x
∗
=Tj,λjx,x
forx∈Eandx∗∈E∗, we haveTj,μjTj,λj=Tj,λjTj,μj=Tj,λj. Puttingμj=λj, we haveTλj=Tλj
and henceTλj is a projection ofEontoFj(G).
As direct consequence of Lemma ., we have the following.
Lemma . Letμjandλjbe invariant means on G.Then Tj,μj=Tj,λj.
Lemma . Assume that an invariant mean exists on G.Then,for each x∈E, the set co{Tj,tx:t∈G} ∩Fj(G)consists of a single point.
Proof Letx∈E andμj be an invariant mean on G. Then we know that Tj,μjx∈Fj(G)
and Tj,μjx∈co{Tj,tx:t∈G}. So, we show thatco{Tj,tx:t∈G} ∩Fj(G) ={Tj,μjx}. Let
x ∈ co{Tj,tx: t∈ G} ∩Fj(G) and > . Then, for x∗ ∈ E∗, there exists an element n
i=αiTj,tixin the setco{Tj,tx:t∈G}such that>C· x∗j· n
i=αiTj,tix–xj. Hence,
we have
>C·x∗j·
n
i=
αiTj,tix–x j
≥sup
t
Tj,tj·
n
i=
αiTj,tix–x j
·x∗j
≥sup
t
n
i=
αiTj,tTj,tix–x j
·x∗j
≥
n
i=
αiTj,tTj,tix–x,x∗
=
n
i=
αiμj,t
Tj,ttix–x,x∗
=μj,t
Tj,tx–x,x∗
=Tj,μjx–x,x
∗.
Since is arbitrary, we haveTj,μjx,x∗=x,x∗for everyx∗∈E∗and henceTj,μjx=x.
4 Ergodic theorems
Now, we can prove mean ergodic theorems for semigroups of linear continuous operators in multi-Banach space.
Theorem . Let{Tj,t:t∈G}be a family of linear continuous operators in a real Banach
space E satisfying Assumption(A).If a net{μα
j :α∈I}of means on G is asymptotically
invariant,i.e.,
μαj –rs∗μαj and μαj –l∗sμαj
converge toin the weak∗topology of mj(G)∗for each s∈G,then there exists a projection Qj
of E on to Fj(G)such that(Q, . . . ,Qj)j≤C,Tj,μα
QjTj,t=Tj,tQj=Qjfor each t∈G,and Qjx∈co{Tj,tx:t∈G}for each x∈E.Furthermore,
the projection Qjonto Fj(G)is the same for all asymptotically invariant nets.
Proof Let μj be a cluster point of net{μαj :α ∈I} in the weak∗ topology of mj(G)∗.
Then μj is an invariant mean on G. Hence, by Lemma ., Tj,μj is a projection of E
ontoFj(G) such that(T,μ, . . . ,Tj,μj)j≤C,Tj,μjTj,t=Tj,tTj,μj=Tj,μj for eacht∈Gand
Tj,μjx∈co{Tj,tx:t∈G}for eachx∈E. SettingQj=Tj,μj, we show thatTj,μαjxconverges
weakly toQjxfor eachx∈E. SinceTj,μαjx∈co{Tj,tx:t∈G}for allα∈Iandco{Tj,tx:t∈G}
is weakly compact, there exists a subnet{Tj,μβ
jx:
β∈J}of{Tj,μαjx:α∈I}such thatTj,μβ
jx
converges weakly to an elementx∈co{Tj,tx:t∈G}. To show thatTj,μα
jxconverges weakly
toQjx, it is sufficient to showx=Qjx. Letx∗∈E∗ands∈G. SinceTj,μβ
jx→xweakly, we
haveμβj,tTj,tx,x∗ → x,x∗andμβj,tTj,tx,Tj,s∗x∗ → x,Tj,s∗x∗=Tj,sx,x∗. On the other
hand, sinceμβj –ls∗μβj → in theweak∗topology, we have
μβj,tTj,tx,x∗
–l∗sμβj,tTj,tx,x∗
=μβj,tTj,tx,x∗
–μβj,tTj,stx,x∗
=μβj,tTj,tx,x∗
–μβj,tTj,tx,Tj,s∗x∗
→.
Hence, we havex,x∗=Tj,sx,x∗and hencex∈Fj(G). So, we obtainQjx=Tj,μjx=x
from Lemma .. That the projectionQjis the same for all asymptotically invariant nets
is obvious from Lemma ..
As direct consequence of Theorem ., we have the following.
Corollary . Let{Tj,t:t∈G}be as in Theorem.and assume that an invariant mean
exists on G.Then there exists a projection Qj of E onto Fj such that(Q, . . . ,Qj)j≤C,
QjTj,t=Tj,tQj=Qjfor each t∈G and Qjx∈co{Tj,tx:t∈G}for each x∈E.
Theorem . Let{Tj,t:t∈G}be as in Theorem..If a net{μαj :α∈I}of means on G
is asymptotically invariant and further μαj –rs∗μαj converges toin the strong topology of mj(G)∗,then there exists a projection Qjof E onto Fj(G)such that(Q, . . . ,Qj)j≤C,
Tj,μαjx converges strongly to Qjx for each x∈E,QjTj,t=Tj,tQj=Qjfor each t∈G,and Qjx∈
co{Tj,tx:t∈G}for each x∈E.
Proof As in the proof of Theorem ., letQj=Tj,μj, whereμjis a cluster point of the net
{μα
j :α∈I}in theweak∗topology ofmj(G)∗. We show thatTj,μαjxconverges strongly to
Qjxfor eachx∈E.
LetE=co{y–Tj,ty:y∈E,t∈G}. Then, for anyz∈E,Tj,μαjzconverges strongly to .
In fact, ifz=y–Tj,sy, then since, for anyy∗∈E∗,
Tj,μαjz,y∗=μαj,t
Tj,t(y–Tj,sy),y∗
=μαj,tTj,ty,y∗
–μαj,tTj,tsy,y∗
≤μα –rs∗μα, . . . ,μαj –r∗sμαj
j·sup t
Tj,ty,y∗
≤μα –rs∗μα, . . . ,μαj –rs∗μαjj·C· yj·y∗j,
we have(T,μαz, . . . ,Tj,μαjz)j≤C· (μα–r∗sμ α
, . . . ,μαj –r∗sμ α
j)j· yj. Using this
inequal-ity, we show thatTj,μαjzconverges strongly to for anyz∈E. Letzbe any element ofE
andbe any positive number. By the definition ofE, there exists an element n
i=ai(yi–
Tj,siyi)in the setco{y–Tj,sy:y∈E,s∈G}such that> C·(z– n
i=ai(yi–T,siyi), . . . ,z– n
i=ai(yi–Tj,siyi))j. On the other hand, from(μα –r∗sμ α , . . . ,μ
α j –rs∗μ
α
j)j→ for all
s∈G, there existsa∈Isuch that, for allα≥αandi= , , . . . ,n,
>μα –rsi∗μα, . . . ,μαj –rsi∗μαjj·Cyij.
This yields
(T,μαz, . . . ,Tj,μαjz)j
≤
T,μα
z–T,μα
n
i=
ai(yi–T,siyi)
,
. . . ,Tj,μαjz–Tj,μαj n
i=
ai(yi–Tj,siyi) j +
T,μα n
i=
ai(yi–T,siyi)
, . . . ,Tj,μαj n
i=
ai(yi–Tj,siyi)
j
≤(T,μα, . . . ,Tj,μαj)j
· z– n i=
ai(yi–T,siyi), . . . ,z– n
i=
ai(yi–Tj,siyi) j + n i=
T,μαj(yi–T,siyi), . . . ,Tj,μαj(yi–Tj,siyi)j
≤C·
z– n i=
ai(yi–T,siyi), . . . ,z– n
i=
ai(yi–Tj,siyi) j +sup i
μα –rsi∗μα, . . . ,μαj –r∗siμαj
j·C· yij
< +
=.
Hence,Tj,μαjZconverges strongly to for eachz∈E.
Next, assume thatx–Tj,μjxfor somex∈E is not contained in the setE. Then, by
the Hahn-Banach theorem, there exists a linear continuous functionaly∗ such thatx– Tj,μjx,y∗= andz,y∗= for allz∈E. So sincex–Tj,tx∈Efor allt∈G, we have
x–Tj,μjx,y
∗=μ
j,t
x–Tj,tx,y∗
= .
This is a contradiction. Hence,x–Tj,μjfor allx∈Eare contained inE. Therefore we find
thatTj,μαjx–Tj,μjx=Tj,μαj(x–Tj,μj) converges strongly to for allx∈E. This completes
By using Theorem ., we can obtain the following corollary.
Corollary . Let E be a real Banach space and Tjbe a linear operator of E into itself
such that exists a constant C with (Tn
, . . . ,Tjn)j≤C for n= , , . . . ,and Tj is weakly
completely continuous,i.e.,Tjmaps the closed unit ball of E into a weakly compact subset
of E.Then there exists a projection Qjof E onto the set Fj(T)of all fixed point of Tjsuch that
(Q, . . . ,Qj)j≤C,the Cesaro means Sj,n=n n
k=Tjkx converges strongly to Qjx for each
x∈E,and TjQj=QjTj=Qj.
Proof Letx∈E. Then, since{Tjnx:n= , , . . .}=Tj({Tn–x:n= , , . . .})⊂Tj(B(,x ·
(c+ ))), whereB(x,r) means the closed ball with centerxand radiusr, the weak closure of{Tn
jx:n= , , . . .}is weakly compact. On the other hand, letG={, , , . . .}with the
discrete topology andμnj be a mean onGsuch thatμnj(fj) = n
i=(
n)fj(i) for eachfj∈mj(G).
Then it is obvious that(μn–rk∗μn, . . . ,μnj –r∗kμnj)j≤kn → for allk∈G. So, it follows
from Theorem . that Corollary . is true.
IfG= [,∞) with the natural topology, then we obtain the corresponding result.
Corollary . Let E be a real Banach space and{Tj,t :t∈[,∞)} be a family of linear
operators of E into itself satisfying Assumption(A).Then there exists a projection Qjof E
onto Fj(G)such that(Q, . . . ,Qj)j≤C, T T
T
j,tx dt converges strongly to Qjx for each
x∈E,and Tj,tQj=QjTj,t=Qjfor each t∈[,∞).
Remark . TTTj,tx dt are weak vector valued integrals with respect to means on G= [,∞). As in Section IV of Rodé [], we can also obtain the strong convergence of the sequences
( –r) ∞
k=
rkTjkx, r→–
and
λ
∞
e–λtTj,tx dt, λ→ + .
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
All authors carried out the proof. All authors conceived of the study, and participated in its design and coordination. All authors read and approved the final manuscript.
Author details
1Department of Mathematics, Science and Research Branch, Islamic Azad University, Ashrafi Esfahani Ave., Tehran, 14778,
Iran.2Department of Mathematics and Computer Science, Iran University of Science and Technology, Tehran, Iran. 3Department of Mathematics Education and the RINS, Gyeongsang National University, Jinju, 660-701, Korea.
Acknowledgements
Yeol Je Cho was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (Grant Number: 2013053358).
References
1. Yosida, K: Mean ergodic theorem in Banach spaces. Proc. Imp. Acad. (Tokyo)14(8), 292-294 (1938)
2. Baillon, J-B: Un théorème de type ergodique pour les contractions non linéaires dans un espace de Hilbert. C. R. Acad. Sci. Paris, Sér. A-B280, A1511-A1514 (1975) (French)
3. Pazy, A: On the asymptotic behavior of iterates of nonexpansive mappings in Hilbert space. Isr. J. Math.26(2), 197-204 (1977)
4. Rodé, G: An ergodic theorem for semigroups of nonexpansive mappings in a Hilbert space. J. Math. Anal. Appl.85(1), 172-178 (1982)
5. Takahashi, W: A nonlinear ergodic theorem for an amenable semigroup of nonexpansive mappings in a Hilbert space. Proc. Am. Math. Soc.81(2), 253-256 (1981)
6. Greenleaf, FP: Invariant Means on Topological Groups and Their Applications. Van Nostrand Mathematical Studies, vol. 16. Van Nostrand-Reinhold, New York (1969)
7. Takahashi, W: Invariant functions for amenable semigroups of positive contractions onL1. Kodai Math. Semin. Rep.
23, 131-143 (1971)
8. Hirano, N, Takahashi, W: Nonlinear ergodic theorems for an amenable semigroup of nonexpansive mappings in a Banach space. Pac. J. Math.112(2), 333-346 (1984)
9. Kido, K, Takahashi, W: Mean ergodic theorems for semigroups of linear operators. J. Math. Anal. Appl.103(2), 387-394 (1984)
10. Dales, HG, Polyakov, ME: Multi-normed spaces. Diss. Math.488, 1-165 (2012)
11. Dales, HG, Moslehian, MS: Stability of mappings on multi-normed spaces. Glasg. Math. J.49(2), 321-332 (2007) 12. Moslehian, MS, Nikodem, K, Popa, D: Asymptotic aspect of the quadratic functional equation in multi-normed
spaces. J. Math. Anal. Appl.355(2), 717-724 (2009)
13. Park, C, Saadati, R: Approximation of a generalized additive mapping in multi-Banach modules and isomorphisms in multi-C∗-algebras: a fixed-point approach. Adv. Differ. Equ.2012, 162 (2012)
14. O’Regan, D, Rassias, JM, Saadati, R: Approximations of ternary Jordan homomorphisms and derivations in multi-C∗
ternary algebras. Acta Math. Hung.134(1-2), 99-114 (2012)
15. Agarwal, RP, Cho, YJ, Park, C, Saadati, R: Approximate homomorphisms and derivation in multi-Banach algebras. Comment. Math.51(1), 23-38 (2011)
16. Kenari, HM, Saadati, R, Cho, YJ: The mean ergodic theorem for nonexpansive mappings in multi-Banach spaces. J. Inequal. Appl.2014, 259 (2014)
17. Day, MM: Amenable semigroups. Ill. J. Math.1, 509-544 (1957)
18. Saluja, GS: Weak convergence theorems for two asymptotically quasi-nonexpansive non-self mappings in uniformly convex Banach spaces. J. Nonlinear Sci. Appl.7(2), 138-149 (2014)
19. Thianwan, T: Convergence criteria of modified Noor iterations with errors for three asymptotically nonexpansive nonself-mappings. J. Nonlinear Sci. Appl.6(3), 181-197 (2013)
20. Purtas, Y, Kiziltunc, H: Weak and strong convergence of an explicit iteration process for an asymptotically quasi-i-nonexpansive mapping in Banach spaces. J. Nonlinear Sci. Appl.5(5), special issue, 403-411 (2012) 10.1186/1029-242X-2014-402
Cite this article as:Kenari et al.:Mean ergodic theorem for semigroups of linear operators in multi-Banach spaces.