The Journal of Fuzzy Mathematics Vol. 19, No. 2, 2011 ?

Los Angeles

A -Summability and Fuzzy Korovkin-type Approximation

George A. Anastassiou

Department of Mathematical Sciences, The University of Memphis, Memphis, TN 38152, USA

E-mail: [email protected]

Kamil Demirci

Sinop University, Faculty of Sciences and Arts, Department of Mathematics, 57000 Sinop, Turkey
E-mail: [email protected]

Sevda Karakus

Sinop University, Faculty of Sciences and Arts, Department of Mathematics, 57000 Sinop, Turkey
E-mail: [email protected]

Abstract:
The aim of this paper is to present a fuzzy Korovkin-type approx-imatiion theorem by
using a matrix summability method. We also study the rates of convergence of fuzzy
using a matrix summability method. We also study the rates of convergence of fuzzy
positive linear operators.
Key words and phrases:
Matrix summability, fuzzy positive linear operators, fuzzy Korovkin type
approximation theorem, fuzzy modulus of continuity, rates of convergence.

1. Introduction
The study of the classical Korovkin type approximation theory is a well-established
area of research, which deals with the problem of approximating a function f by means
of a sequence {Ln (f )} of positive linear operators. Most of the classical approximation
operators tend to converge to the value of the function being approximated. However, at
points of discontinuity, they often converge to the average of the left and right limits of
the function. In this case, the matrix summability methods of the Cesaro type are
applicable to correct the lack of convergence [7]. The main purpose of using
summability theory has always been to make a nonconvergent sequence converge. Some
results regarding matrix summability for positive linear operators may be found in the
paper [4], [5], [19]. In this paper using a A -summation process we study the
__________________________
Received July, 2009
1066-8950/11 $8.50
© 2011 International Fuzzy Mathematics Institute
Los Angeles
2 George A. Anastassiou, Kamil Demirci and Sevda Karakus

approximation properties of sequence of positive linear operators defined on the space of

all fuzzy continuous functions on the interval [a, b ] .
Now, we recall some basic definitions and results used in the paper.
A fuzzy number is a function m : Æ [0,1] , which is normal, convex, uppersemi-
continuous and the closure of the set supp ( m ) is compact, where supp ( m )  {x Π:
m (x) > 0} . Denote the set of all fuzzy numbers are denoted by F . Let

[ m ]0 ¬ {x Œ : m (x) > 0} and [ m ] ¬ {x Œ : m (z ) ≥ r} , 0 < r £ 1 .

r

Then, it is well-known [11] that, for each r Œ[0,1] , the set [ m ] is a closed and bounded
r

interval of . For any u, v ΠF and l Π, it is possible to define uniquely the sum

u ≈ v and the product u v as follows:

[u ≈ v]r = [u]r + [v]r and [ l u ] = l [u ] , 0 £ r £ 1 .

r r

Denote the interval [u] by ÈÎ u-( ) , u+( ) ˘˚ , where u-( ) £ u+( ) and u-( ) , u+( ) Œ for r Œ[0,1] .
r r r r r r r

Then, for u, v ΠF , define

u ≺ v ¤ u-( ) £ v-( ) and u+( ) £ v+( ) for all 0 £ r £ 1 .

r r r r

Define the following metric D : F ¥ F Æ + by

{
D (u, v ) ¬ sup max u-(r) - v-(r ) , u+(r ) - v+(r) .
r Œ[0,1]
}
We have that D is a metric on F . Then ( F , D ) is a complete metric space, (see
[8]), with the properties
D ( u Šw , v Šw ) = D ( u , v ) , " u, v , w ΠF ,
D (k u, k v ) = k D ( u , v ) , " u, v ΠF and "k Π,
D ( u ≈ v , w ≈ e ) £ D ( u, w ) + D ( v , e ) , " u, v , w, e Œ F .

Let f , g : [a, b ] Æ F be fuzzy number valued functions. Then, the distance between
f and g on is given by

{
D* (f , g ) ¬ sup sup max f-(r ) - g-(r) , f+(r) - g+(r) .
xŒ r Œ[0,1]
}
Anastassiou and Duman [3] introduced the fuzzy analog of A -statistical convergence
by using the metric D .
Let ( mn )nŒN be a fuzzy number valued sequence and let A = ajn is a non-negative ( )
regular summability matrix. Recall that the regularity conditions on a matrix A are
know as Silverman-Toeplitz conditions in the functional analysis (see for details, [12]).
Then, ( mn )nŒN is A -statistically convergent to m Œ F , which is denoted by
stA -limD ( mn , m ) = 0 if for every e > 0 .
n Æ•
 A -Summability and Fuzzy Korovkin-type Approximation 3

lim Â
j Æ• n:D ( m , m ) ≥ e
ajn = 0
n

holds. Some basic results regarding A -statistical convergence for number sequences
may be found in the papers [9], [13], [16]. Of course, the case of A = C1 , the Cesaro
matrix of order one, immediately reduces to the statistical convergence of fuzzy valued
sequences which defined by Nuray and Savas [18]. Also, if A is replaced by the identity
matrix, then we have the fuzzy convergence introduced by Mathoka [15].
Let A ¬ An ( ) , A n = akjn
n ≥1
( )be a sequence of infinite non-negative real
k, jŒ

matrixces. For a sequence of real numbers, x = xj ( ) jŒ

, the double sequence

{
Ax ¬ (Ax)k : k, n Œ
n
}
defined by (Ax)k ¬ S •j =1akjn xj is called the A -transform of x whenever the series
n

converges for all k and n . A sequence x is said to be A -summable to L if

•
lim  akjn xj = L
k Æ• j =1

uniformly in n ([6], [20]).

If An = A for some matrix A , then A -summability is the ordinary matrix summa-
bility by A . If, akjn = 1 k , for n £ j £ k + n , (n = 1, 2, ) , and akjn = 0 otherwise, then
A -summability reduces to almost convergence [14].

2. A fuzzy korovkin type theorem

In this section we prove a fuzzy Korovkin-type approximation theorem via the concept
of A -summation process.
Let f : [a, b] Æ F be fuzzy number valued functions. Then f is said to be fuzzy
continuous at x0 Œ[a, b ] provided that whenever xn Æ x0 , then D f ( xn ) , f (x0 ) Æ 0 ( )
as n Æ • . Also, we say that f is fuzzy continuous on [a, b] if it is fuzzy continuous at
every point x Œ[a, b ] . The set all fuzzy continuous functions on the interval [a, b] is
denoted by C F [a, b ] (see, for instance, [2]). Notice that C F [a, b ] is only a cone not a
vector space. Now let L : C F [a, b ] Æ C F [a, b] be an operator. Then L is said to be
fuzzy linear if, for every l1 , l2 Œ , f1 , f2 Œ C F [a, b] , and x Œ[a, b ] , L ( l1 f1 ≈ l2
f2 ; x) = l1 L (f1 ; x) ≈ l2 L (f2 ; x) holds. Also L is called fuzzy positive linear
operator if it is fuzzy linear and the condition L (f ; x ) ≺ L ( g; x ) is satisfied for any
f , g Œ C F [a, b] and all x Œ[a, b ] with f (x) ≺ g (x) .
Throughout the paper the symbol f denotes the usual supremum norm of f .
4 George A. Anastassiou, Kamil Demirci and Sevda Karakus

{ } of positive linear operators of C [a, b] into itself is called an A -

A sequence Lj
summation process on C [a, b ] if {L (f )} is A -summable of f for every f ΠC [a, b ] ,
j

i.e.,
•
lim  akjn Lj (f ) - f = 0 , uniformly in n ,
k Æ• j =1
(1)

where it is assumed that the series in (1) converges for each k , n and f .
{ } be sequence of positive linear operators of C [a, b] into itself such that for
Let Lj
each k, n Œ
•
sup  akjn Lj (1) < • . (2)
n , k j =1

•
Furthermore, for k, n Œ and f Œ C [a, b ] , let Bkn (f ; x) ¬  akjn Lj (f ) which is well
j =1

defined by (2) and belongs to C [a, b ] .

In this paper, we establish a theorem of Korovkin type with respect to the convergence
behavior (1) for a sequence of positive linear operators of C F [a, b ] into itself. So our
•
results are extensions of type (1) lim  akjn Lj (f ) - f = 0 , uniformly in n . This answers
k Æ• j =1

the question of Professor F . Altomare of University of Bari, Italy, to the affirmative.

{ }
If a sequence Lj of positive linear operators of C F [a, b] into itself is called an A -
summation process on C F [a, b ] if {L (f )}
j is A -summable to f for every
Ê• ˆ
f Œ C F [a, b ] ; i.e., lim D* Á  akjn Lj (f ) , f ˜ = 0 , uniformly in n .
k Æ• Ë j =1 ¯
Throughout the paper we use the test functions ei (x) = xi (i = 0,1, 2) .
Then Anastassiou [2] obtained the following fuzzy Korovkin theorem.
Theorem 1 [2]. Let {Ln }nΠbe a sequence of fuzzy positive linear operators from
C F [a, b ] into itself. Assume that there exists a corresponding sequence Lj { } nŒ
of
positive linear operators from C [a, b] into itself with the property

{L (f ; x)}( ) = L (f ( ) ; x)
r r
n ± j ± (3)

for all x Œ[a, b ] , r Œ[a, b] , n Œ and f Œ C F [a, b ] . Assume further that lim L (ei )
n Æ• j

- ei = 0 for each i = 0,1, 2 . Then, for all f ΠC F [a, b ] , we have lim D (Ln (f ) , f ) = 0 .
*
n Æ•

Recently the A -statistical analog of Theorem 1 has been studied by Anastassiou and
Duman [3]. It will be read as follows.

( )
Theorem 2 [3]. Let A = ajn be a non-negative regular summability matrix and let
{Ln }nΠbe a sequence of fuzzy positive linear operators from C F [a, b] into itself.
 A -Summability and Fuzzy Korovkin-type Approximation 5

Assume that there exists a corresponding sequence Lj { } nŒ

of positive linear operators
from C [a, b] into itself with the property (3). Assume further that stA - lim L (ei ) -
n Æ• j

ei = 0 for each i = 0,1, 2 . Then, for all f ΠC F [a, b ] , we have stA - lim D (Ln (f ) , f ) *
n Æ•

= 0.
We now give the following generalization by using a matrix summability method.

Theorem 3. Let A = An ( ) n ≥1
be a sequence of infinite non-negative real matrices
such that
•
sup  akjn < • (4)
n ,k j =1

and Lj { } jŒ
be a sequence of fuzzy positive linear operators from C F [a, b] into itself.

Assume that there exists a corresponding sequence Lj { } jŒ

of positive linear operators
from C [a, b] into itself with the property (3) and satisfying (2). Assume further that

•
lim  akjn Lj (ei ) - ei = 0 for each i = 0,1, 2
k Æ• j =1
(5)

Ê• ˆ
uniformly in n . Then, for all f Œ C F [a, b ] , we have lim D* Á  akjn Lj (f ) , f ˜ = 0
k Æ• Ë j =1 ¯
uniformly in n .

Proof. Let f Œ C F [a, b ] , x Œ[a, b ] and r Œ[0,1] . By the hypothesis, since f±(r) Œ
C [a, b ] , we may write, for every e > 0 , that there exists a number d > 0 such that
f±( ) (y ) - f±( ) (x) < e holds for every y Œ[a, b ] satisfying y - x < d . Then we get for
r r

all y Œ[a, b] , that

r (y - x )
2

f±( ) (y ) - f±( ) (x) £ e + 2M ±( )

r r
, (6)
2
d

where M ±( ) ¬ f±( ) . Now using the linearity and the positivity of the operators Lj
r r

( ) (
• •
and considering inequality (6), we can write  akjn Lj f±(r) ; x -f±(r ) (x) £  akjn Lj f±(r ) (y )
j =1 j =1

(r ) ( y - x )
Ê ˆ
(x) ; x) + f± (x) Â a
2
• •
- f± (r ) (r ) n
Lj (1; x) - 1 £ Â a Lj Á x + 2M ±
n
; x ˜ + f±(r)
d
kj kj 2
j =1 j =1 Ë ¯
2M ±(r )
( )
• • •
 akjn Lj (1; x) - 1 £ e +e  akjn Lj (1; x) - 1 +  akjn Lj (y - x) ; x + M ±( )
2 r

j =1 j =1 d 2
j =1

2M ±(r )
( ) ( )
• • •
 akjn Lj (1; x) - 1 £ e + e + M ±(r )  akjn Lj (1; x) - 1 +  akjn Lj (y - x) ; x =
2

j =1 j =1 d 2
j =1
6 George A. Anastassiou, Kamil Demirci and Sevda Karakus

2M ±(r ) È • n
( ) ( )
• •
= e + e + M ±( ) Â akjn Lj (1; x) - 1 +
r
 akj Lj y 2 ; x -2x  akjn Lj (y; x) +
j =1 d 2 ÍÎ j =1 j =1

(r )
˘
( ) Âa 2M ± È • n
( )
• •
x2 Â akjn Lj (1; x)˙ £ e + e + M ±( ) Lj (1; x) - 1 + Í Â akj Lj y ; x - x
r n 2 2

d 2 Î j =1
kj
j =1 ˚ j =1

˘
( ) Âa
• • •
+2 x  akjn Lj (y; x) - x + x2  akjn Lj (1; x) - 1 ˙ £ e + e + M ±(r) n
kj Lj (1; x) - 1 +
j =1 j =1 ˚ j =1

2M ±(r ) 4M ±(r ) c 2M ±(r ) c 2

 akjn Lj (y 2 ; x) - x2
• • •
+ Â akjn Lj (y; x) -x + Â akjn Lj (1; x) - 1
d 2
j =1 d 2
j =1 d 2
j =1

Ê (r ) (r )
2M ± c ˆ 2 • 4M ± c •
£ e + Á e + M ±(r ) + Â akjn Lj (e0 ; x) - e0 + Â akjn Lj (e1 ; x) - e1 +
Ë d 2 ˜¯ j =1 d2 j =1

{
(r ) •
 akjn Lj (e2 ; x) - e2 where c ¬ max { a , b } . If we take K ±( )(e ) ¬ max e +M ±( )
2M ± r r

d 2
j =1

2c 2 M ±(r ) 4cM ±(r ) 2M ±(r ) Ô̧ Ï

 akjn Lj (f±( ) ; x) -f±( )(x) £ e +K ±( ) (e )Ì Â akjn
• •
+ , , ˝ ,then we get
r r r

d2 d2 d 2 Ô˛ j =1 Ó j =1
• • ¸
Lj (e0 ; x) - e0 + Â akjn Lj (e1 ; x) - e1 + Â akjn Lj (e2 ; x) - e2 ˝ . Then we observe that
j =1 j =1 ˛

( ) Ï •
•
sup max  akjn Lj f±( ) ; x - f±( ) (x) £ e + sup max K ±(r) (e ) Ì Â akjn Lj (e0 ; x) - e0 +
r r

r Œ[0,1] j =1 r Œ[0,1] Ó j =1
¸
 akjn Lj (e1 ; x) - e1 +  akjn Lj (e2 ; x) - e2 ˝ . Taking K ¬ K (e ) ¬ sup max {K -( ) (e ) ,
• •
r

j =1 j =1 ˛ r Œ[0,1]

} ( Ï
)
• •
K +(r) (e ) , we get sup sup max  akjn Lj f±( ) ; x - f±( ) (x) £ e + K (e ) Ì sup  akjn
r r

xŒ[a , b ] r Œ[0,1] j =1 ÓxŒ[a ,b] j =1

• • ¸
Lj (e0 ; x) - e0 + sup  akjn Lj (e1 ; x) - e1 + sup  akjn Lj (e2 ; x) - e2 ˝ Consequently, we
xŒ[a , b ] j =1 xŒ[a , b ] j =1 ˛
Ê • ˆ • •
get D* Á Â akjn Lj (f ; x) , f (x)˜ £ e + K Â akjn Lj (e0 ) - e0 + K Â akjn Lj (e1 ) - e1 +
Ë j =1 ¯ j =1 j =1

•
K Â akjn Lj (e2 ) - e2 .
j =1

By taking limit as k Æ • and by using (5), we obtain the desired result.

If we take A n = I , the identity matrix and An = A , for some matrix A in Theorem 3,
we immediately get Theorem 1 and Theorem 2, respectively.
Remarks. We now present three examples of sequences of positive linear operators.
The first one shows that our approximation theorem works but statistical version of
Theorem 2 does not work. The second one shows that our approximation theorem works
but Theorem 1 does not work and the last one shows that statistical version of Theorem 2
works but Theorem 3 does not work.
 A -Summability and Fuzzy Korovkin-type Approximation 7

Assume that A ¬ A n ( ) n ≥1
( )
= akjn
k , jŒ
is a sequence of infinite matrices defined by
n
akj = 1 k if n £ j £ k + n , (n = 1, 2, ), and akjn = 0 otherwise, then A -summability
reduces to almost convergence.
(i) Take uj = ( ) {(-1) } . j
Observe that u is almost convergent to zero, but it is not
statistically convergent. The consider the Fuzzy Bernstein-type polynomials as follows:
j Ê jˆ
Ê kˆ
Bj( ) (f ; x) = 1 + uj
F
(k =0 Ë
)
≈ Á ˜ xk (1 - x)
k¯
j -k
f Á ˜ , where f Œ C F [0,1] , x Œ[0,1] and
Ë j¯

{ } ( ) ( )  ÊÁË kˆ˜¯ x (1 - x) Ê kˆ
(r ) j j
. So we write Bj(F ) (f ; x) = Bj f±(r) ; x = 1 + uj f±(r) Á ˜ ,
j -k
jΠk
± k =0 Ë j¯
where f± Œ C [0,1] . Then, we get Bj (e0 ; x) = 1 + uj
(r )
( ) ( )
Bj (e1 ; x) = 1 + uj x Bj (e2 ; x)

( )(
= 1 + uj x + x - x 2
(( 2
) j )) Since u is almost convergent to zero we get
•
lim  akjn Bj (ei ) - ei = 0 for each i = 0,1, 2 uniformly in n . So, by Theorem 3, we
k Æ• j =1

Ê• ˆ
obtain, for all f Œ C F [0,1] , that lim D* Á  akjn Bj (f ) , f ˜ = 0 uniformly in n . However,
k Æ• Ë j =1 ¯
since u is not statistically convergent to zero, we observe that Bj { } satisfies the
Theorem 3 but it does not satisfy the statistical version of Theorem 2.
( )
(ii) Let vj = (1, 0,1, 0, ) . We see that vj is almost convergent to 1 2 , but not ( )
j Ê jˆ
Ê1 ˆ Ê kˆ
Then consider L(jF ) (f ; x) = Á + vj ˜ ≈ Á ˜ xk (1 - x)
j -k
convergent. f Á ˜ , for
Ë2 ¯ k = 0 Ë k¯ Ë j¯
f Œ C F [0,1] , x Œ[0,1] and j Œ . Hence Lj { } satisfies the Theorem 3 but it does not
satisfy the Theorem 1.
(iii) Now consider the sequence of 0’s and 1’s defined as follows: 0, , 0,1, ,1, 0, ,
Æ100 ¨ Æ10 ¨
0,1, ,1,1, . where the blocks of 0’s are increasing by factors of 100 and the blocks of
1’s are increasing by factors of 10. if we denote this sequence by yj , then yj is not ( ) ( )
almost convergent but it is statistically convergent to zero [17]. Then we define
j Ê jˆ
Ê kˆ
(
Tj(F ) (f ; x) = 1 + y j )
≈ Á ˜ xk (1 - x)
k = 0 Ë k¯
j -k
f Á ˜ for f Œ C F [0,1] , x Œ[0,1] and
Ë j¯
jŒ . We observe that Tj { } jŒ
satisfies the statistical version of Theorem 2 but it does
not satisfy the Theorem 3.

3. Rate of convergence
In this section we study the rates of convergence of the sequence of fuzzy positive
linear operators examined in Theorem 3.
8 George A. Anastassiou, Kamil Demirci and Sevda Karakus

Let f : [a, b] Æ F . Then the (first) fuzzy modulus of continuity of f , which is

introduced by Gal [10] 9see also [2]), is defined by w1(F ) (f , d ) ¬ sup D (f ( x ) ,
x , y Œ[a , b ]; x - y £d

f (y )) for any 0 < d £ b - a . As in [1], we have w1(F ) (f , d ) = sup max w1 f-(r ) , d ,

r Œ[0,1]
{ ( )
(
w1 f+(r ) , d )} . Then we give the following.
( ) be a sequence of infinite non-negative real matrices
Theorem 4. Let A = A n
n ≥1
•
such that sup  a < • and {L } n
be a sequence of fuzzy positive linear operators
kj j jŒ
n ,k j =1

from C [a, b] into itself. Assume that there exists a corresponding sequence {L }
F of j jŒ

positive linear operators from C [a, b ] into itself with the property (3) and satisfying (2).
Suppose that Lj satisfy the following conditions:
•
(i) lim  akjn Lj (e0 ) - e0 = 0 uniformly in n
k Æ• j =1

( )
•
(ii) lim w1(F ) f , mkn  akjn Lj (e0 ) - e0 = 0 uniformly in n
k Æ• j =1

( )
•
(iii) lim w1( ) f , mkn = 0 uniformly in n , where mkn = Â akjn Lj (j ) and j (y ) =
F
k Æ• j =1

Ê• ˆ
for each x Œ[a, b ] . Then for all f Œ C F [a, b ] we have lim D* Á  akjn Lj (f ) , f ˜
(y - x ) 2 k Æ• Ë j =1 ¯
= 0 uniformly in n .
Proof. Using the linearity and the positivity of the operators Lj we may write, for each

 akjn Lj (f±( ) ; x) - f±( ) (x) (

• • •
jŒ £ f±(r ) (x)  akjn Lj (1; x)- 1 + akjn Lj f±(r) (y )
r r
, that
j =1 j =1 j =1

Ê Ê r d (y - x ) ˆ ˆ
)
• •
- f±(r ) (x ) ; x £ M ±( ) Â akj Lj (e0 ; x) - e0 + Â akjn Lj Á w1 Á f±( ) , £ M ±(r)
r n

d ˜¯ ; x˜
j =1 j =1 Ë Ë ¯
ÊÊ È y - x ˘ˆ ˆ
( )
• • •
 akjn Lj (e0 ; x) - e0 +  akjn Lj Á Á1 + Í (r )
˙ ˜ w1 f± , d ; x˜ £ M ±(r) Â akjn Lj (e0 ; x)
j =1 j =1 ËË Î d ˚¯ ¯ j =1

ÊÊ y-x ˆ ˆ
( Â akjn Lj (e0 ; x) - e0 + w1 (f±( ) , d )
2

)
• •
- e0 + Â akjn Lj Á Á1 + (r )
˜ w1 f± , d ; x˜ £ M ±(r) r

j =1 ÁË Ë d 2
¯ ˜¯ j =1

w1 f , d ( (r )
) Âa
( ) Â akjn Lj (e0 ; x) - e0 + w1 (f±( ) ,
• • •
 akjn Lj (e0 ; x) +
±
Lj (y - x) ; x £ M ±(r)
n 2 r

d 2 kj
j =1 j =1 j =1

(
w1 f , d (r )
) Âa
d ) Â akjn Lj (e0 ; x) - e0 ( ) ( )
• ± •
+ w1 f±(r) , d + Lj (y - x) ; x
n 2
where
d 2 kj
j =1 j =1
 A -Summability and Fuzzy Korovkin-type Approximation 9

 akjn Lj (f±( ) ; x) - f±( ) (x) £ M ±( )  akjn Lj (e0 ; x)

• •
M ±( ) ¬ f±( ) . Then we observe that
r r r r r

j =1 j =1

( (r )
w1 f± , d ) Âa
( ) Âa ( )
• •
- e0 +w1 f±(r ) , d n
Lj (e0 ; x) - e0 + w1 f±(r ) , d + n
Lj (j ; x) If we
d2
kj kj
j =1 j =1

( )
• •
take d = mkn , we conclude that  akjn Lj f±( ) ; x - f±( ) (x) £ M ±( )  akjn Lj (e0 ; x) - e0
r r r

j =1 j =1

( ) Âa ( ) Ê• ˆ
•
+ w1 f±(r) , mkn n
Lj (e0 ; x) - e0 + 2w1 f±(r ) , mkn . Then we get D* Á Â akjn Lj (f ) , f ˜ £
j =1
kj
Ë j =1 ¯

( ) ( )
• •
M Â akjn Lj (e0 ) - e0 + Â akjn Lj (e0 ) - e0 w1( ) f , mkn +2w1(F ) f , mkn
F
where M ¬
j =1 j =1

{
sup M -r , M +r .
r Œ[0,1]
} Letting k Æ • and using the hypothesis (i), (ii) and (iii) we get

Ê• ˆ
lim D* Á Â akjn Lj (f ) , f ˜ = 0 uniformly in n , which completes the proof.
k Æ• Ë j =1 ¯
We note that the rate of convergence of the involved operators may be obtained from
(7).

References
[1] G. A. Anastassiou, Fuzzy approximation by fuzzy Convolution type operators, Computers and
Mathematics with Applications, 48 (2004), 1369-1386.
[2] G. A. Anastassiou, On basic fuzzy korovkin theory, Studia Univ. Babes-Bolyai Math., 50 (2005), 3-10.
[3] G. A. Anastassiou and O. Duman, statistical fuzzy approximation by fuzzy positive linear operators,
Comput. Math. Appl., 55 (2008), 573-580.
[4] O. G. Athhan and C. Orhan, Matrix summability and positive linear operators, Positivity, 11 (2007),
387-389.
[5] O. G. Athhan and C. Orhan, Summation process of positive linear operators, Computers and
Mathematics with Applications, 56 (2008), 1188-1195.
[6] H. T. Bell, Order summability and almost convergence, Proc. Am. Math. Soc., 38 (1973), 548-553.
[7] R. Bojanic and M. K. Khan, Summability of hermite-fejer interpolation for functions of bounded
variation, J. Nat. Sci. Math., 32 No.1 (1992), 5.
[8] Wu Congxin and Ma Ming, Embedding problem of fuzzy number space, Part 1, fuzzy Sets and
Systems, 44 (1991), 33-38-10.
[9] A. R. Freedman and J. J. Sember, Densities and summability, Pacific J. Math., 95 (1981), 293-305.
[10] S. G. Gal, Approximation theory in fuzzy setting, In: Handbook of analytic-computational methods in
applied mathematics, Chapman and Hall/CRC, Boca Raton, FL, (2000), 617-666.
[11] R. J. Goetschel and W. Voxman, Elementary fuzzy calculus, Fuzzy Sets and Systems, 18 (1986), 31-43.
[12] G. H. Hardy, Divergent series, Oxford Uni. Press, London, (1949).
[13] E. Kolk, Matrix summability of statistically convergent sequences, Analysis, 13 (1993), 77-83.
[14] G. G. Lorentz, A contribution to the theory of divergent sequences, Acta Math., 80 (1948), 167-190.
[15] M. Matloka, Sequences of fuzzy numbers, BUSEFAL, 28 (1986), 28-37.
[16] H. I. Miller, A measure theoretical subsequence characterization of statistical convergence, Trans.
Amer. Math. Soc., 347 (1995), 1811-1819.
[17] H. I. Miller and C. Orhan, On almost convergent and statistically convergent subsequences, Acta Math.
Hungar., 93 (2001), 135-151.
10 George A. Anastassiou, Kamil Demirci and Sevda Karakus

[18] F. Nuray and E. Savas, Statistical convergence of sequences of fuzzy numbers, Math. Slovaca, 45
(1995), 269-273.
[19] C. Radu, A -summability and approximation of continuous periodic functions, Studia Univ. Babes-
Bolyai Math., 52 (2007), 155-161.
[20] M. Steiglitz,Eine verallgemeinerung des begriffs der fastkonvergenz, Math. Japon., 18 (1973), 53-70.