Equi-Statistical -Convergence of Positive

Linear Operators

Sevda Karakuş and Kamil Demirci

Sinop University, Faculty of Sciences and Arts, Department of Mathematics,
57000 Sinop, Turkey

Abstract

In this paper we introduce the notion of equi-statistical convergence which is

stronger than the uniform convergence (in ordinary sense), statistical uniform con-
vergence and statistical uniform convergence. Then, we also give its use in the
Korovkin-type approximation theory. We also compute the rate of equi-statistical
convergence of sequence of positive linear operators.

Key words: Statistical uniform convergence, equi-statistical convergence,

statistical pointwise convergence, positive linear operator, Korovkin-type
approximation theorem.

1 Introduction

The concept of convergence of a sequence of real numbers has been extended

to statistical convergence by Fast [7]. Let K be the subset of positive integers.
Then the natural density of K is given by

1
lim jKn j
(K) = n!1
n
if it exists where Kn = fk 2 K : k ng and the vertical bars denote the car-
dinality of the set. A sequence x = fxk g of numbers is statistically convergent
to L provided that, for every " > 0, fk : jxk Lj "g = 0 hold. In this case
we write st lim x = L:

Recently various kind of statistical convergence which is stronger than the

statistical convergence has been introduced by Mursaleen and Edely [11]. We
…rst recall this convergence method.

Preprint submitted to Elsevier Science 21 April 2010

Let be a one-to-one mapping from the set N into itself. A continuous linear
functional ' de…ned on the space l1 of all bounded sequences is called an
invariant mean (or mean) [13] if and only if

(i) '(x) 0 when the sequence x = fxk g has xk 0 for all k;

(ii) '(e) = 1, where e = (1; 1; :::);

(iii) '(x) = '((x (n) )) for all x 2 l1 :

Thus, mean extends the limit functional on c of all convergent sequences in

the sense that '(x) = lim x for all x 2 c [12]. Consequently, c V where V
is the set of bounded sequences all of whose means are equal. It is known
[14] that

V = x 2 l1 : limtpm (x) = L uniformly in m, L = lim x

p

where
xm + x (m) + x 2 (m) + ::: + x p (m)
tpm (x) := :
p+1

We say that a bounded sequence x = fxk g is convergent if and only if

x2V .

Let
( )
s
V = x 2 l1 : st lim tpm (x) = L uniformly in m, L = lim x :
p

A sequence x = fxk g is said to be statistically convergent to L if and only

if x 2 V s (see, for details, [11]). In this case we write ( ) lim xk = L. That
is,
1
lim jfp n : jtpm (x) Lj "gj = 0, uniformly in m:
n n

Note that, if x is statistically convergent then x is statistically convergent.

However, its converse is not always true. Such an example was given in [11].

Let f and fn belong to C (X), which is the space of all continuous real valued
functions on a compact subset X of the real numbers. The usual supremum
norm on the spaces C (X) is given by

kf kC(X) := sup jf (x)j for f 2 C (X) :

x2X

2
In [3] a kind of convergence for sequence lying between pointwise and uniform
statistical convergence is presented. We …rst recall these convergence methods.

De…nition 1 [3]ffn g is said to be statistically pointwise convergent to f on

X if st limn!1 fn (x) = f (x) for each x 2 X, i.e., for every " > 0 and for
each x 2 X;
jfk n : jfk (x) f (x)j "gj
lim = 0:
n!1 n
In this case, we denote this limit by fn ! f (stat) on X.

De…nition 2 [3]ffn g is said to be equi-statistically convergent to f on X if

for every " > 0,
!
jfk n : jfk (x) f (x)j "gj
lim sup = 0:
n!1 x2X n

In this case, we denote this limit by fn ! f (equi stat) on X.

De…nition 3 [3]ffn g is said to be statistically uniform convergent to f on X

if st limn!1 kfn f kC(X) = 0, or
n o
k n : kfk f kC(X) "
lim = 0:
n!1 n
In this case, we denote this limit by fn f (stat) on X.

Now we give the following de…nitions:

De…nition 4 fe (x) := ffn (x)g is said to be statistically pointwise convergence

to f on X if

st lim tpm fe (x) = f (x)

n!1
uniformly in m

for each x 2 X; i.e. for every " > 0 and for each x 2 X
n o
p n : tpm fe (x) f (x) "
lim =0 uniformly in m
n!1 n
where
fm (x) + f (m) (x) + f 2 (m) (x) + ::: + f p (m) (x)
tpm fe (x) := :
p+1

Then, it is denoted by fn ! f (stat ( )) on X.

De…nition 5 fe (x) := ffn (x)g is said to be equi-statistically convergence

3
to f on X if for every " > 0
0 n o 1
p n : tpm fe (x) f (x) "
lim @sup A=0 uniformly in m:
n!1 x2X n

This limit is denoted by fn ! f (equi stat ( )) on X.

De…nition 6 fe (x) := ffn (x)g is said to be statistically uniform convergence

to f on X if for every " > 0

p n : tpm fe (x) f (x) "

C(X)
lim =0 uniformly in m:
n!1 n
This limit is denoted by fn f (stat ( )) on X.

Using the above de…nitions, the next result follows immediately.

Lemma 7 fn f on X (in ordinary sense) implies fn f (stat) on X,

which also implies fn f (stat ( )) on X. Furthermore fn f (stat ( ))
on X implies fn ! f (equi stat ( )) on X and fn ! f (equi stat ( )) on
X implies fn ! f (stat ( )) on X. Also, fn ! f on X (in ordinary sense)
implies fn ! f (stat) on X, which also implies fn ! f (stat ( )) on X and
fn ! f (equi stat) on X implies fn ! f (equi stat ( )) on X.

However one can construct an example which guarantees that the converses
of Lemma 7 are not always true. Such an example is in the following:

Example 8 Consider the case (n) = n + 1. Let g (x) = 0, x 2 [0; 1] and

de…ne continuous functions gk : [0; 1] ! R, k 2 N, by the formula
8 h i
> 2
>
> 22(k+1) x 1
; if x 2 1
; 1 1
;
>
> 2k 2k 2k 1 2k+1
>
>
>
>
>
>
>
< h i
2
1 1 1
gk (x) =
>2
2(k+1)
x 2k 1
; if x 2 2k 1 2k+1
; 2k1 1 (1)
>
>
>
>
>
>
>
>
>
>
>
: 0 ; otherwise.

Then observe that gk ! g = 0 (equi stat ( )) on [0; 1], however fgk g is

not statistically uniform convergent to the function g = 0 on the interval
[0; 1]. Also, fgk g is not uniform convergent (in ordinary sense) and statistically
uniform convergent to the function g = 0 on [0; 1] :

For a sequence fLn g of positive linear operators on C (X) Korovkin [10] …rst
introduced the su¢ cient conditions for the uniform convergence of Ln (f ) to a

4
function f by using the test function ei de…ned by ei (x) = xi (i = 0; 1; 2). Later
many researchers investigate these conditions for various operators de…ned on
di¤erent spaces. Furthermore, in recent years, with the help of the concept of
uniform statistical convergence, which is a regular (non-matrix) summability
transformation, various statistical approximation results have been proved ([1],
[2], [4], [5], [6], [8], [9]). Then, it was demonstrated that those results are more
powerful than the classical Korovkin theorem.

In this paper, we obtain the Korovkin-type approximation theorem by means

of the concept of equi-statistical convergence in De…nition 5.

2 A Korovkin-Type Approximation Theorem

In this section using equi-statistical convergence, we prove the following

Korovkin-type approximation theorem.

Theorem 9 Let X be a compact subset of the real numbers and let L := fLn g
be a sequence of positive linear operators acting from C (X) into itself. Then
for all f 2 C (X),

Ln (f ) ! f (equi stat ( )) on X (2)

if and only if
Ln (ei ) ! ei (equi stat ( )) on X (3)
i
with ei (x) = x , i = 0; 1; 2:

PROOF. Since each ei 2 C (X), i = 0; 1; 2, the implication (2))(3) is obvi-

ous.

Now suppose that (3) holds. Let f 2 C (X) and x 2 X be …xed. By the
continuity of the point x, we may write that for every " > 0, there exists a
number > 0 such that jf (y) f (x)j < " for all y 2 X satisfying jy xj < .
Since

jf (y) f (x)j = jf (y) f (x)j X (y) + jf (y) f (x)j XnX (y)

where X = [x ; x + ] \ X and A denotes the characteristic function of

the set A. Then we have
(y x)2
jf (y) f (x)j " + 2M 2 (4)

for all y 2 X, where M := kf kC(X) . Using the positivity and monotonicity of

Ln , we obtain from (4)

5
 jtpm (L (f ; x)) f (x)j

Lm (f ; x) + L (m) (f ; x)
+ ::: + L p (m) (f ; x)
= f (x)
p+1
Lm (f ; x) + L (m) (f ; x)
+ ::: + L p (m) (f ; x)
=
p+1
Lm (e0 ; x) + L (m) (e0 ; x)
+ ::: + L p (m) (e0 ; x)
f (x)
p+1
Lm (e0 ; x) + L (m) (e0 ; x)
+ ::: + L p (m) (e0 ; x)
+ f (x) f (x)
p+1
Lm (jf (y) f (x)j ; x) + L (m) (jf (y) f (x)j ; x) + ::: + L p (m) (jf (y) f (x)j ; x)
p+1
+ jf (x)j jtpm (L (e0 ; x)) e0 (x)j

2M
"+ "+M + 2 ke2 kC(X) jtpm (L (e0 ; x)) e0 (x)j

4M 2M
+ 2 ke1 kC(X) jtpm (L (e1 ; x)) e1 (x)j + 2 jtpm (L (e2 ; x)) e2 (x)j :

Then, we get

jtpm (L (f ; x)) f (x)j " + K fjtpm (L (e0 ; x)) e0 (x)j (5)

+ jtpm (L (e1 ; x)) e1 (x)j

+ jtpm (L (e2 ; x)) e2 (x)jg

where K := " + M + 2M2 ke2 kC(X) + 2 ke1 kC(X) + 1 : Now, for a given r > 0,
choose " > 0 such that " < r. Then, it follows from (5) that

jfp n : jtpm (L (f ; x)) f (x)j rgj

r "
p n : jtpm (L (e0 ; x)) e0 (x)j
3K
r "
+ p n : jtpm (L (e1 ; x)) e1 (x)j
3K
r "
+ p n : jtpm (L (e2 ; x)) e2 (x)j
3K
which gives

6
 jfp n : jtpm (L (f ; x)) f (x)j rgj
sup (6)
x2X n
n o
r "
p n : jtpm (L (e0 ; x)) e0 (x)j 3K
sup
x2X n
n o
r "
p n : jtpm (L (e1 ; x)) e1 (x)j 3K
+sup
x2X n
n o
r "
p n : jtpm (L (e2 ; x)) e2 (x)j 3K
+sup :
x2X n

Then using the hypothesis (3) and considering De…nition 5 in (6),

!
jfp n : jtpm (L (f ; x)) f (x)j rgj
lim sup = 0 uniformly in m
n!1 x2X n

for every r > 0. The proof is complete.

Remark 10 Now we show that our result Theorem 9 is stronger than classical
Korovkin theorem and statistical Korovkin theorem.

Let X = [0; 1] and consider the following Bernstein-Kantorovich operators:

n
! (k+1)=(n+1)
Z
X n n k
k
Un (f ; x) = (n + 1) x (1 x) f (t) dt
k=0 k
k=(n+1)

where x 2 [0; 1], f 2 C [0; 1] and n 2 N. Using these polynomials, we introduce

the following positive linear operators on C [0; 1] :

Tn (f ; x) = (1 + gn (x)) Un (f ; x) ; x 2 [0; 1] ; f 2 C [0; 1] (7)

where gn (x) is given by (1). Then observe that

Tn (e0 ; x) = (1 + gn (x)) e0 (x)

!
n 1
Tn (e1 ; x) = (1 + gn (x)) e1 (x) +
n + 1 2 (n + 1)
" #
n (n 1) 2nx 1
Tn (e2 ; x) = (1 + gn (x)) e2 (x) + + :
(n + 1)2 (n + 1)2 3 (n + 1)2

where ei (x) = xi , i = 0; 1; 2. Consider the case (n) = n + 1. Since gn ! g =

0 (equi stat ( )) on [0; 1], we conclude that

Tn (ei ) ! ei (equi stat ( )) on [0; 1] for each i = 0; 1; 2

7
So fTn g satisfy all hypothesis of Theorem 9 and we immediately see that

Tn (f ) ! f (equi stat ( )) on [0; 1] for all f 2 C [0; 1] :

However, since fgn g is not uniform convergent (in ordinary sense) to the
function g = 0 on [0; 1], we can say that Tn (ei ) is not uniform convergent
to ei for i = 0; 1; 2. Also since fgn g is not statistically uniform convergent
to the function g = 0 on [0; 1], we can say that Tn (ei ) is not statistically
uniform convergent to ei on [0; 1] for i = 0; 1; 2. So the classical Korovkin
theorem and statistical Korovin theorem do not work for our operators de…ned
by (7). Therefore, this application clearly shows that our Theorem 9 is a non-
trivial generalization of the classical and the statistical case of the Korovkin-
type results introduced in [10] and [8], respectively.

3 Rate of Equi-Statistical Convergence

In this section, we study the rates of equi-statistical convergence of positive

linear operators de…ned on C (X) with the help of modulus of continuity.

De…nition 11 A sequence fe (x) := ffn (x)g is equi-statistical convergent

to a function f with the rate of 2 (0; 1) if for every " > 0
0 n o 1
p n : tpm fe (x) f (x) "
lim @sup A=0 uniformly in m
n!1 x2X n1

where
fm (x) + f (m) (x) + f 2 (m) (x) + ::: + f p (m) (x)
tpm fe (x) := :
p+1

In this case, it is denoted by

fn f = o(n ) (equi stat ( )) on X:

Using this de…nition, we obtain the following auxiliary result.

Lemma 12 Let ffn g and fgn g be function sequences belonging to C (X). As-
sume that fn f = o(n 1 ) (equi stat ( )) on X and gn g = o(n 2 ) (equi stat ( ))
on X: Let := min f 1 ; 2 g : Then we have

(i) (fn f ) (gn g) = o(n ) (equi stat ( )) on X,

(ii) (fn f ) = o(n 1 ) (equi stat ( )) on X, for any real number .

8
PROOF. (i) Assume that fn f = o(n 1 ) (equi stat ( )) on X and gn
g = o(n 2 ) (equi stat ( )) on X: Then, since := min f 1 ; 2 g for " > 0,
observe that

n o
p n: tpm fe (x) f (x) (tpm (ge (x)) g (x)) "
(8)
n1
n o n o
p n: tpm fe (x) f (x) "
2
+ p n : jtpm (ge (x)) g (x)j "
2
n1
n o n o
p n: tpm fe (x) f (x) "
2
p n : jtpm (ge (x)) g (x)j "
2
+
n1 1 n1 2

Now, by taking the limit as n ! 1 in (8) and using the hypotheses, we

conclude that
0 n o 1
p n: tpm fe (x) f (x) (tpm (ge (x)) g (x)) "
lim @sup A = 0; uniformly in m,
n!1 x2X n1

which completes the proof of (i). Since the proof of (ii) is similar, we omit it.

Now we recall that the modulus of continuity of a function f 2 C (X) is

de…ned by

! (f ; ) = sup jf (y) f (x)j ( > 0) :

jy xj ;x;y2X

It is also well know that, for any > 0 and for all f 2 C (X)

! (f ; ) (1 + [ ]) ! (f ; )

where [ ] is de…ned to be the greatest integer less than or equal to :

Then we have the following result.

Theorem 13 Let L := fLn g be a sequence of positive linear operators acting

from C (X) into itself. Assume that the following conditions hold:

Ln (e0 ) e0 = o(n 1 ) (equi stat ( )) on X (9)

and
! f; pm = o(n 2 ) (equi stat ( )) on X (10)
q
where pm := tpm (L ('; x)) with ' (y) = (y x)2 : Then we have for all
f 2 C (X)
Ln (f ) f = o(n ) (equi stat ( )) on X
where := min f 1 ; 2 g :

9
PROOF. Let f 2 C (X) and x 2 X be …xed. Using the properties of ! and
the positivity and monotonicity of Ln , we get, for any > 0 and n 2 N; that

jtpm (L (f ; x)) f (x)j

Lm (f ; x) + L (m) (f ; x)
+ ::: + L p (m) (f ; x)
= f (x)
p+1
Lm (jf (y) f (x)j ; x) + L (m) (jf (y) f (x)j ; x) + ::: + L p (m) (jf (y) f (x)j ; x)
p+1
+ jf (x)j jtpm (L (e0 ; x)) e0 (x)j

jy xj jy xj jy xj
Lm (! f ; ; x) + L (m) (! f; ; x) + ::: + L p (m) (! f ; ; x)
p+1
+ jf (x)j jtpm (L (e0 ; x)) e0 (x)j
h i h i h i
jy xj jy xj jy xj
Lm (1 + ; x) + L (m) (1 + ; x) + ::: + L p (m) (1 + ; x)
! (f ; )
p+1
+ jf (x)j jtpm (L (e0 ; x)) e0 (x)j

(y x)2 (y x)2 (y x)2

Lm (1 + 2 ; x) + L (m) (1 + ; x) + ::: + L
2 p (m) (1 + 2 ; x)
! (f ; )
p+1
+ jf (x)j jtpm (L (e0 ; x)) e0 (x)j

! (f ; ) + jtpm (L (e0 ; x)) e0 (x)j ! (f ; )

Lm ('; x) + L (m) ('; x)

+ ::: + L p (m) ('; x) ! (f ; )
+ 2
p+1
+ jf (x)j jtpm (L (e0 ; x)) e0 (x)j

! (f ; )
! (f ; ) + jtpm (L (e0 ; x)) e0 (x)j ! (f ; ) + 2 tpm (L('; x))

+ jf (x)j jtpm (L (e0 ; x)) e0 (x)j :

Then we have

jtpm (L (f ; x)) f (x)j 2! (f ; ) + jtpm (L (e0 ; x)) e0 (x)j ! (f ; ) (11)

+M jtpm (L (e0 ; x)) e0 (x)j

10
 q
where := pm := tpm (L ('; x)) and M := kf kC(X) . Hence, given " > 0, it
follows from (11) and Lemma 12 that

jfp n : jtpm (L (f ; x)) f (x)j "gj

(12)
n1
n o n q o
" "
p n : ! (f ; ) 6
p n : jtpm (L (e0 ; x)) e0 (x)j 3
+
n1 2 n1 1

n q o n o
" "
p n : ! (f ; ) 3 p n : jtpm (L (e0 ; x)) e0 (x)j 3M
+ +
n1 2 n1 1

where := min f 1; 2 g. Letting n ! 1 in (12), we conclude from (9) and

(10) that
!
jfp n : jtpm (L (f ; x)) f (x)j "gj
lim sup = 0; uniformly in m;
n!1 x2X n1

which means

Ln (f ) f = o(n ) (equi stat ( )) on X:

References

[1] G.A. Anastassiou, O. Duman, A Baskakov type generalization of statistical

Korovkin theory, J. Math. Anal. Appl. 340 (2008) 476-486.

[2] G.A. Anastassiou, O. Duman, Statistical fuzzy approximation by fuzzy positive

linear operators, Comput. Math. Appl. 55 (2008) 573-580.

[3] M. Balcerzak, K. Dems, A. Komisarski, Statistical convergence and ideal

convergence for sequences of functions, J. Math. Anal. Appl. 328 (2007) 715-729.

[4] O. Duman, M.K. Khan, C. Orhan, A-Statistical convergence of approximating

operators, Math. Inequal. Appl. 4 (2003) 689-699.

[5] O. Duman, E. Erkuş, V. Gupta, Statistical rates on the multivariate

approximation theory, Math. Comput. Modelling 44 (2006) 763-770.

[6] E. Erkuş, O. Duman, H.M. Srivastava, Statistical approximation of certain

positive linear operators constructed by means of the Chan-Chyan-Srivastava
polynomials, Appl. Math. Comput. 182 (2006) 213-222.

[7] H. Fast, Sur la convergence statistique, Colloq. Math. 2 (1951) 241-244.

11
[8] A.D. Gadjiev, C. Orhan, Some approximation theorems via statistical
convergence, Rocky Mountain J. Math. 32 (2002), 129-138.

[9] S. Karakuş, K. Demirci, O. Duman, Equi-statistical convergence of positive

linear operators, J. Math. Anal. Appl. 339 (2008) 1065-1072.

[10] P.P. Korovkin, Linear Operators and Approximation Theory, Hindustan Publ.
Co., Delhi, 1960.

[11] M. Mursaleen, O.H.H. Edely, On the invariant mean and statistical convergence,
Applied Mathematics Letters 22 (2009) 1700 1704.

[12] M. Mursaleen, On some new invariant matrix methods of summability, Q. J.

Math. 34 (1983) 77–86.

[13] R.A. Raimi, Invariant means and invariant matrix methods of summability,
Duke Math. J. 30 (1963) 81-94.

[14] P. Schaefer, In…nite matrices and invariant means, Proc. Amer. Math. Soc. 36
(1972) 104–110.

E-mail addresses: [email protected] (S. Karakuş)

[email protected] (K. Demirci)

12