• Nenhum resultado encontrado

A NOTE ON HADAMARD INEQUALITIES FOR THE PRODUCT OF THE CONVEX FUNCTIONS

N/A
N/A
Protected

Academic year: 2017

Share "A NOTE ON HADAMARD INEQUALITIES FOR THE PRODUCT OF THE CONVEX FUNCTIONS"

Copied!
3
0
0

Texto

(1)

IJRRAS 5 (2) ● November 2010 Amrahov ● A Note On Hadamard Inequalities

168

A NOTE ON HADAMARD INEQUALITIES FOR THE PRODUCT

OF THE CONVEX FUNCTIONS

Şahin Emrah Amrahov

The Computer Engineering Department of Engineering Faculty, Ankara University, Keçiören, Ankara, Turkey.

ABSTRACT

The main aim of the present note is to prove new Hadamard like integral inequalities for the product of the convex functions.

Keywords: integral inequality, convex functions, Hadamard inequality

Mathematics subject classification (2000):26D15, 26D99

1. INTRODUCTION

Let

f

be a real valued convex function defined on the interval

[

a

,

b

]

. Then

2

)

(

)

(

)

(

1

2

b

f

a

f

dx

x

f

a

b

b

a

f

b

a

 

(1)

is known as Hadamard’s inequality for the convex function. [1] If

f

uv

and

u

,

v

are convex functions then we have

2

)

(

)

(

)

(

)

(

)

(

)

(

1

2

2

b

v

b

u

a

v

a

u

dx

x

v

x

u

a

b

b

a

v

b

a

u

b

a

 

 

(2)

by Hadamard inequality.

Therefore by Cauchy-Scwartz inequality

2

)

(

)

(

)

(

)

(

2

)

(

)

(

)

(

)

(

a

v

a

u

b

v

b

u

2

a

u

2

b

v

2

a

v

2

b

u

(3)

the inequality

2

)

(

)

(

)

(

)

(

)

(

)

(

1

u

2

a

u

2

b

v

2

a

v

2

b

dx

x

v

x

u

a

b

b

a

(4)

holds.

Note that if

u

,

v

are convex functions, then the function

f

uv

may not be convex function. Example1.

u

(

x

)

x

2

,

v

(

x

)

(

1

x

)

2 are convex functions defined on

[

0

,

1

]

,but the function

2 2 2

2

(

1

)

(

)

)

(

)

(

)

(

x

u

x

v

x

x

x

x

x

f

is not convex function on this interval.

In fact, the first and second derivatives

f

(

x

)

,

f



(

x

)

of

f

(

x

)

can be calculated by the formulas.

)

1

2

)(

(

2

)

(

2

x

x

x

x

f

)

(

4

)

1

2

(

2

)

(

x

x

2

x

2

x

f



Therefore we have

1

2

1



f

.

It means that the function

f

uv

is not convex function on the interval

[

0

,

1

]

.

Example2. For the convex functions

u

(

x

)

x

2

,

v

(

x

)

(

1

x

)

2 defined on

[

0

,

1

]

we have

30

1

)

(

)

(

)

(

1

1

0

2

2

a

u

x

v

x

dx

x

x

dx

b

b

a

(2)

IJRRAS 5 (2) ● November 2010 Amrahov ● A Note On Hadamard Inequalities

169

.

0

)

1

(

)

1

(

,

0

)

0

(

)

0

(

v

u

v

u

Hence the Hadamard inequality

0

2

)

1

(

)

1

(

)

0

(

)

0

(

30

1

)

(

)

(

)

(

1

1

0

2

2

v

u

v

u

dx

x

x

dx

x

v

x

u

a

b

b

a

for the non-convex function

f

uv

does not hold.. On the other hand, since

0

)

1

(

,

1

)

0

(

,

1

)

1

(

,

0

)

0

(

2 2 2 2

v

v

u

u

the inequality

2

1

2

)

1

(

)

0

(

)

1

(

)

0

(

30

1

)

(

)

(

)

(

1

1 2 2 2 2

0

2

2

v

v

u

u

dx

x

x

dx

x

v

x

u

a

b

b

a

holds. It means that although the function

f

uv

is no convex function we have that the inequality (4) is true for these functions. Our aim is to investigate the inequality (4) when

f

uv

is non-convex function.

2. MAIN RESULTS

Our main result is the following theorem.

Theorem. Let

u

and

v

are nonnegative convex functions defined on the interval

[

a

,

b

]

. Then the inequality (4) holds.

To prove of the theorem we need the following lemma.

Lemma. Let

u

is a nonnegative convex function defined on the interval

[

a

,

b

]

. Then the function

u

2 is also convex function on the interval

[

a

,

b

]

Proof. For arbitrary

x

,

y

[

a

,

b

]

and

k

[

0

,

1

]

we have

0

)

(

)

(

)

(

2

)

(

0

))

(

)

(

(

2 2

2

y

u

y

u

x

u

x

u

y

u

x

u

)

(

)

(

)

(

)

(

2

u

x

u

y

u

2

x

u

2

y

(5) Multiplying both sides of the inequality (5) by

k

(

1

k

)

we get

)

(

)

1

(

)

(

)

1

(

)

(

)

(

)

1

(

2

k

k

u

x

u

y

k

k

u

2

x

k

k

u

2

y

Therefore

)

(

]

)

1

(

)

1

[(

)

(

)

(

)

(

)

(

)

1

(

2

k

k

u

x

u

y

k

k

2

u

2

x

k

k

2

u

2

y

So

)

(

)

1

(

)

(

)

1

(

)

(

)

(

)

(

)

(

)

1

(

2

k

k

u

x

u

y

ku

2

x

k

2

u

2

x

k

u

2

y

k

2

u

2

y

Hence

)

(

)

1

(

)

(

)

(

)

1

(

)

(

)

(

)

1

(

2

)

(

2 2 2 2

2 2

y

u

k

x

ku

y

u

k

y

u

x

u

k

k

x

u

k

Therefore

)

(

)

1

(

)

(

)]

(

)

1

(

)

(

[

ku

x

k

u

y

2

ku

2

x

k

u

2

y

(6) Since

u

(

x

)

is a nonnegative convex function we have

)

(

)

1

(

)

(

)

)

1

(

(

kx

k

y

ku

x

k

u

y

u

2 2

(

kx

(

1

k

)

y

)

[

ku

(

x

)

(

1

k

)

u

(

y

)]

u

(7)

(3)

IJRRAS 5 (2) ● November 2010 Amrahov ● A Note On Hadamard Inequalities

170

)

(

)

1

(

)

(

)

)

1

(

(

2 2

2

kx

k

y

ku

x

k

u

y

u

(8)

The inequality (8) proves that the function

u

2 is a convex function. Proof of the theorem.

By the lemma the functions

u

2 and

v

2 are convex functions. By the Hadamard inequality for these functions we

have

2

)

(

)

(

)

(

1

2

u

2

a

u

2

b

dx

x

u

a

b

b

a

(9)

2

)

(

)

(

)

(

1

v

2

x

dx

v

2

a

v

2

b

a

b

b

a

(10)

Multiplying the inequalities (9) and (10) we get

2

)

(

)

(

2

)

(

)

(

)

(

)

(

)

(

1

2 2 2 2 2 2

2

b

v

a

v

b

u

a

u

dx

x

v

dx

x

u

a

b

b

a

b

a

(11)

By Cauchy-Scwartz inequality

b

a

b

a b

a

dx

x

v

dx

x

u

dx

x

v

x

u

(

)

(

)

2

(

)

2

(

)

2

(12)

Hence by (11) and (12) we get

2

)

(

)

(

2

)

(

)

(

)

(

)

(

)

(

1

2 2 2 2

2

2

b

v

a

v

b

u

a

u

dx

x

v

x

u

a

b

b

a

The last inequality means that the inequality (4) holds.

Example3. Prove the inequality

sin

8

2

10

2

dx

x

x

Solution. Since the functions

u

(

x

)

sin

x

8

and

x

x

v

(

)

1

are nonnegative convex

functions on

,

2

we have by inequality (4)

2

4

1

1

)

8

2

(sin

)

8

(sin

8

sin

1

2 2

2 2

2

dx

x

x

Hence

4

5

8

8

8

sin

1

2

2

2

dx

x

x

Therefore

10

2

8

sin

2

dx

x

x

Note that

x

x

x

f

(

)

sin

8

is not a convex function on

,

2

. 3. REFERENCES

[1] J.E.Pecaric, F.Proschan and Y.L.Tong, “Convex Functions, Partial Orderings and Statistical

Referências

Documentos relacionados

A infestação da praga foi medida mediante a contagem de castanhas com orificio de saída do adulto, aberto pela larva no final do seu desenvolvimento, na parte distal da castanha,

i) A condutividade da matriz vítrea diminui com o aumento do tempo de tratamento térmico (Fig.. 241 pequena quantidade de cristais existentes na amostra já provoca um efeito

didático e resolva as ​listas de exercícios (disponíveis no ​Classroom​) referentes às obras de Carlos Drummond de Andrade, João Guimarães Rosa, Machado de Assis,

A empresa vencedora poderá retirar o instrumento equivalente (Nota de Empenho) na Reitoria do IFRJ à Rua Pereira de Almeida, 88 - Praça da Bandeira - Rio de Janeiro/RJ. A

A GA5 manteve a linha do tempo no rodapé da página, com o conteúdo separado por frases de acordo com cada coluna, mas com sangria; com o título da obra no topo

O cuidador familiar e o idoso em sua totalidade aprendem a conviver com a condição de dependência, com dinâmicas sociais construídas nos significados do cuidar,

Despercebido: não visto, não notado, não observado, ignorado.. Não me passou despercebido

Tem estudantes que manejam redes sociais, controlam blogs, mas quando são questionados sobre conceitos ou sobre como usar os elementos do blog, das redes, de forma mais efetiva,