American Journal of Engineering Research (AJER) 2014

w w w . a j e r . o r g

Page 27
American Journal of Engineering Research (AJER)
e-ISSN : 2320-0847 p-ISSN : 2320-0936
Volume-03, Issue-02, pp-27-29
www.ajer.org

Research Paper Open Access


Remarks on one S.S. Dragomirs Result

Katerina Anevska, Risto Maleski
Faculty of informatics, FON University, Macedonia
Faculty of informatics, FON University, Macedonia

Abstract: - In theorem 1 [1], S.S. Dragomir gave bounds for the normalized Jensen functional defined by
convex function f , which one is defined on strictly convex subset C of vector space X . Further, using
inequality (2.1) of normed space ( ,|| ||) X he proved the inequalities (3.1), (3.2) and (3.3), and after that from
inequality (3.3) he performed inequality (3.6), which was previously proved in [2]. In this paper well give an
example, which shows that inequalities (3.3) are not correct and will show how the inequality (3.2) implies (3.6)
given in [1].
2010 Mathematics Subject Classification. Primary 46B20, 46B99; Secondary 51M16.

Keywords: - Convex function, Jensens inequality, triangle inequality

I. INTRODUCTION
Let X be a vector space, C convex subset of X ,
n
P set of all nonnegative n tuples
1 2
( , ,..., )
n
p p p such that
1
1
n
i
i
p
=
=

, : f C R a convex function,
1 2
( , ,..., )
n
x x x C = e x ,
n
P e p and
1 1
( , , ) ( ) ( ) 0
n n
n i i i i
i i
J f p f x f p x
= =
= >

x p (1)
be the normalized Jensen functional. In [1], for functional (1), S.S. Dragomir gave elementary proof of the
following theorem (theorem of bounds for the normalized Jensen functional ).

Theorem 1. If ,
n
P e p q , 0
i
q > , for each 1, 2,..., i n = then
1 1
( , , ) max { } ( , , ) ( , , ) min { }
i i
i i
p p
n n n
q q
i n i n
J f J f J f
s s s s
> > x q x p x q . (2)

Furthermore, using the fact, that in normed space ( ,|| ||), X the function : ,
p
f X R ( ) || || , 1
p
p
f x x p = > is
convex on X , S.S. Dragomir proved that inequality (2) implies the following inequalities
1
1 1 1 1
1
1 1 1 1
[ || || || || ] max { } || || || || ,
|| || || || [ || || || || ] min { },
i
i
i
i
n n n n
p
p p p p
j j j j j j j j
q
i n
j j j j
n n n n
p
p p p p
j j j j j j j j
q
i n
j j j j
q x q x p x p x
p x p x q x q x
s s
= = = =
s s
= = = =
>
>


(3)
And letting
1
j
n
q = , for 1, 2,..., j n = he get the following inequalities
American Journal of Engineering Research (AJER) 2014


w w w . a j e r . o r g

Page 28
1
1
1 1 1 1
1
1
1 1 1 1
[ || || || || ] max{ } || || || || ,
|| || || || [ || || || || ] min { }.
n n n n
p p p p p
j j i j j j j
i n
j j j j
n n n n
p p p p p
j j j j j j i
i n
j j j j
x n x p p x p x
p x p x x n x p

s s
= = = =

s s
= = = =
>
>


(4)
Finally, letting
1
|| ||
i
i
x
p = , for \ {0}, 1, 2,...,
i
x X i n e = and also using the inequalities (4) S.S. Dragomir get the
following :
1 1
|| ||
1
1 1 1 1
1 1
|| ||
1
1 1 1 1
[ || || || || ] max{|| ||} || || || || ,
|| || || || [ || || || || ] min {|| ||},
j
j
j
j
n n n n
x
p p p p p
j i j j
x
i n
j j j j
n n n n
x
p p p p p
j j j i
x
i n
j j j j
x x x n x
x n x x x

s s
= = = =

s s
= = = =
>
>


(5)
By which for 1 p = he get the following
|| ||
1
1 1 1
|| ||
1
1 1 1
[ || ||] max{|| ||} || || || ||,
|| || || || [ || ||] min {|| ||},
j
j
j
j
n n n
x
i j j
x
i n
j j j
n n n
x
j j i
x
i n
j j j
n x x x
x x n x
s s
= = =
s s
= = =
>
>


(6)
proved in [2], and their generalization was given by Mitani, Saito, Kato and Tamura, [3] and also by Peari and
Raji, [4].

II. MAIN COMMENT
Example 1. Let
n
X = R , 2 n > and || || be an Euclids norm. Then the vertex (0,..., 0,1, 0,..., 0),
i
i
x =
1, 2,..., i n = satisfy the following
|| ||
|| || 1,
i
i
x
i i
x
x x = = , for 1, 2,..., i n = and
|| ||
1 1
|| || || ||
i
i
n n
x
i
x
i i
x n
= =
= =

.
According to this, for 1 p > the inequalities (5) applies the following ones
2 2 2
1
p p p
p
n n n n n n n

> > ,
So, we get that for 2 n > and 1 p > is true that
1
1
p
n

= , and that is contradiction.

At first, it seemed that inequalities (3) - (6) get correct by (2). So this procedure [2] is citated by L. Maligranda.
However, the example 1 shows that inequality (5) is not correct if 1 p > . The error occurred in a choice of
numbers
1
|| ||
i
i
x
p = , \ {0},
i
x X e 1, 2,..., i n = . In fact, according to Theorem 1 these numbers have to satisfy the
condition
1
1
n
i
i
p
=
=

. The mentioned condition is not satisfied for arbitrary vectors \ {0},

i
x X e 1, 2,..., i n =
and for thus selected numbers , 1, 2,...,
i
p i n =

Anyway, Theorem 1, i.e. inequality (4) implies (6).
Let 0
i
o > ,for 1, 2,..., i n = , and
1
, 1, 2,...,
i
n
k
k
i
p i n
o
o
=
= =

. So,
n
P e p and If we take that
1
, 1, 2,...,
i
n
k
k
i
p i n
o
o
=
= =

in (4), we get the following inequalities:
1 1
1
1 1 1 1 1
1 1
1
1 1 1 1 1
[ || || || || ] max{ } || || ( ) || || ,
|| || ( ) || || [ || || || || ] min { }.
n n n n n
p p p p p p
i i i i i i i i
i n
i i i i i
n n n n n
p p p p p p
i i i i i i i i
i n
i i i i i
x n x x x
x x x n x
o o o o
o o o o

s s
= = = = =

s s
= = = = =
>
>


(7)
American Journal of Engineering Research (AJER) 2014


w w w . a j e r . o r g

Page 29
Letting
1
|| ||
i
i
x
o = , for \ {0}, 1, 2,...,
i
x X i n e = , in inequalities (7) we get the following:
1 1 1 1
|| || || ||
1
1 1 1 1 1
1 1 1 1
|| || || ||
1
1 1 1 1 1
|| || || || [ || || ( ) || || ] min {|| ||},
[ || || ( ) || || ] max {|| ||} || || || || .
i
i i
i
i i
n n n n n
x
p p p p p p
i i i i
x x
i n
i i i i i
n n n n n
x
p p p p p p
i i i i
x x
i n
i i i i i
x n x x x
x x x n x

s s
= = = = =

s s
= = = = =
>
>



Finally, by using the inequalities above for 1 p = , we get inequalities (6).

Remarks. In the end, we can note that for || ||
i i
x o = , , 1, 2,...,
i
x X i n e = , inequality (7) implies the following
inequalities
1 1 1
1
1 1 1 1 1
1 1 1
1
1 1 1 1 1
[ || || || || ] max{|| ||} || || ( || ||) || || || || ,
|| || ( || ||) || || || || [ || || || || ] min {|| ||},
n n n n n
p p p p p p
i i i i i i i
i n
i i i i i
n n n n n
p p p p p p
i i i i i i i
i n
i i i i i
x n x x x x x x
x x x x x n x x
+
s s
= = = = =
+
s s
= = = = =
>
>



In which, for 1 p = we get the inequalities below:
2
1
1 1 1 1
2
1
1 1 1 1
[ || || || ||] max {|| ||} || || || || || ||,
|| || || || || || [ || || || ||] min {|| ||}.
n n n n
i i i i i i
i n
i i i i
n n n n
i i i i i i
i n
i i i i
x x x x x x
x x x x x x
s s
= = = =
s s
= = = =
>
>



Similarly, as (6), if the vertex , 1, 2,...,
i
x X i n e = are such that || || 1
i
x = , then they become equalities.

REFERENCES
[1] S. S. Dragomir, Bounds for the normalized Jensen functional, Bull. Austral. Math. Soc. 74, No. 3, 2006,
pp. 471-478
[2] Kato, M., Saito, K.S., Tamura, T. Sharp triangle inequality and its reverse in Banach spaces,
Mathematical Inequalities & Application, Vol. 10, No. 2, 2007, pp. 17-31
[3] K. I. Mitani, K. S. Saito, M. Kato, T. Tamura, On sharp triangle inequalities in Banach spaces, J. Math.
Anal. Appl. 336, 2007, pp. 1178-1186
[4] J. E. Peari, R. Raji, The Dunkl-Williams inequality with n elements in normed linear spaces, Math.
Inequa. Appl., No. 2, 2007, pp. 461-470
[5] L. Maligranda, Some remarks on the triangle inequality for norms, Banach J. Math, Anal., 2008, No. 2,
pp. 31-41