volume 7, issue 5, article 159, 2006.
Received 09 June, 2006;
accepted 08 December, 2006.
Communicated by:G. Bennett
Abstract Contents
JJ II
J I
Home Page Go Back
Close Quit
Journal of Inequalities in Pure and Applied Mathematics
AN INEQUALITY BETWEEN COMPOSITIONS OF WEIGHTED ARITHMETIC AND GEOMETRIC MEANS
FINBARR HOLLAND
Mathematics Department University College Cork, Ireland
EMail:f.holland@ucc.ie
c
2000Victoria University ISSN (electronic): 1443-5756 165-06
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page2of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
http://jipam.vu.edu.au
Abstract
LetPdenote the collection of positive sequences defined onN. Fixw∈P. Let s, t, respectively, be the sequences of partial sums of the infinite seriesP
wk andP
sk, respectively. Givenx ∈P, define the sequencesA(x)andG(x)of weighted arithmetic and geometric means ofxby
An(x) =
n
X
k=1
wk
sn xk, Gn(x) =
n
Y
k=1
xwkk/sn, n= 1,2, . . .
Under the assumption thatlogtis concave, it is proved thatA(G(x))≤G(A(x)) for allx∈P, with equality if and only ifxis a constant sequence.
2000 Mathematics Subject Classification:Primary 26D15
Key words: Weighted averages, Carleman’s inequality, Convexity, Induction.
Contents
1 Introduction. . . 3 2 The Main Result . . . 7
References
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page3of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
1. Introduction
In [13], Kedlaya proved the following theorem.
Theorem 1.1. Letx1, x2, . . . , xn, w1, w2, . . . , wnbe positive real numbers, and definesi =w1+w2 +· · ·+wi, i= 1,2, . . . , n. Assume that
(1.1) w1
s1
≥ w2 s2
≥ · · · ≥ wn sn
. Then
(1.2)
n
Y
i=1 i
X
j=1
wj si xj
!wi/sn
≥
n
X
j=1
wj sn
j
Y
i=1
xwii/sj,
with equality if and only ifx1 =x2 =· · ·=xn.
Choosingwto be a constant sequence, we recover the inequality
(1.3) n
v u u t
n
Y
i=1
1 i
i
X
j=1
xj
!
≥ 1 n
n
X
j=1
j
v u u t
j
Y
i=1
xi,
which Kedlaya [12] had previously established, thereby confirming a conjecture of the author [9]. The strict inequality prevails in (1.3) unlessx1 =x2 =· · ·= xn. Evidently, inequality (1.3) is a sharp refinement of Carleman’s well-known one [4, 7]. (Indeed, as a tribute to Carleman, the author was led to formulate (1.3) in an attempt to design a suitable problem for the IMO when it was held
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page4of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
http://jipam.vu.edu.au
in Sweden in 1991. However, unbeknownst to him at the time, two stronger versions of it had already been stated, without proof, by Nanjundiah [17].)
In passing, we note that (1.3) is also a simple consequence of more general results found by Bennett [2,3], and Mond and Peˇcari´c [16].
Also in [13], Kedlaya deduced a weighted version of Carleman’s inequality from Theorem1.1, viz.,
Theorem 1.2. Letw1, w2, . . . be a sequence of positive real numbers, and define si =w1+w2+· · ·+wi, fori= 1,2, . . .. Assume that
(1.4) w1
s1 ≥ w2
s2 ≥ · · · .
Then, for any sequencea1, a2, . . . of positive real numbers withP
kwkak<∞,
∞
X
k=1
wkaw11/sk· · ·awkk/sk < e
∞
X
k=1
wkak.
Carleman’s classical inequality is obtained from this by settingwi = 1, i = 1,2, . . .. This beautiful result has attracted the attention of many authors, and has been proved in a variety of ways. It has also been extended in different directions by a host of people. Anyone interested in knowing the history of Carleman’s inequality, and such matters, is urged to consult [11], which has an extensive bibliography. In addition, the fascinating monograph by Bennett [1]
contains some very interesting developments of it, and mentions, inter alia, the significant extensions of it made by Cochran and Lee [5], Heinig [8] and Love [14,15]. Readers interested in its continuous analogues should also read [18].
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page5of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
Kedlaya expressed a doubt that the monotonicity condition (1.4) was needed in Theorem1.2. His suspicions were well-founded, for, already in 1925, Hardy [6,7], following a suggestion made to him by Pólya, proved this statement with- out any extra hypothesis on the weights. In fact, in the presence of condition (1.4), a much stronger conclusion can be drawn, as the author has recently dis- covered [10]. This begs the question: does Theorem 1.1 also hold under less stringent conditions on the weights than (1.1)? It is trivially true whenn = 1, and a convexity argument shows it also holds without any restriction on the weights when n = 2. However, as Kedlaya himself pointed out, the result is false in general. As he mentions, a necessary condition for the truth of Theorem 1.1is that
wn sn
sn−1
≤ w1
s1 w1
w2 s2
w2
· · ·
wn−1 sn−1
wn−1
.
On the other hand, examples show that the sufficient assumption (1.1) is not necessary. For instance, with n = 3, w1 = 2, w2 = 1, w3 = 3, then w2/s2 <
w3/s3, so that condition (1.1) fails, yet 2a+√3
a2b+ 3√6 a2bc3
6 ≤ 6
s a2
2a+b 3
2a+b+ 3c 6
3
,
for all a, b, c > 0, with equality if and only if a = b = c. (This is a simple consequence of the fact that, if
F(x, y) = (2 +x+ 3√ xy)6 (2 +x3)(2 +x3+ 3y2)3,
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page6of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
http://jipam.vu.edu.au
then
maxx≥0 max
y≥0 F(x, y) = max
x≥0
"
1 2 +x3
maxy≥0
(2 +x+ 3√ xy)2 2 +x3+ 3y2
3#
= max
x≥0
(4 + 10x+x2+ 3x4)3 (2 +x3)4
= 72,
which can be verified in a routine manner, even by non-calculus arguments.
Alternatively, it can be inferred as a special case of Theorem2.1which follows.
Moreover, there is equality if and only ifx=y= 1.)
As an examination of his proof of Theorem 1.1 reveals, Kedlaya actually proved something stronger than (1.2) under the hypothesis (1.1), namely, de- noting byLn, Rnthe left-hand and right-hand sides of (1.2), then
(1.5)
L1 R1
s1
≤ L2
R2 s2
≤ · · · ≤ Ln
Rn sn
.
However, this statement is false in general, and, in particular, is not implied by (1.2). To see this, note that, withn = 3, and the same choice of weightsw1 = 2, w2 = 1, w3 = 3as before, so that (1.2) holds, the claim that (L3/R3)s3 ≥ (L2/R2)s2 is equivalent to the statement that
2(2a+b+ 3c)(2a+ 3
√
a2b)≥(2a+ 3
√
a2b+ 36
√
a2bc3)2, ∀a, b, c >0.
However, this is not true generally, as may be seen by takinga= 1, b= 64, c = 121. So, Kedlaya proved a stronger statement with the hypothesis that the se- quence si/wi is increasing. By adopting a different proof-strategy, we show here that (1.2) holds under a weaker hypothesis than this.
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page7of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
2. The Main Result
The purpose of this note is to present the following result which strengthens Theorem1.1.
Theorem 2.1. Letx1, x2, . . . , xn, w1, w2, . . . , wnbe positive real numbers. De- finesi =w1+w2+· · ·+wi, i= 1,2, . . . , n. Assume that
(2.1) s2k
wk+1 ≥
k−1
X
j=1
sj, k = 2,3, . . . , n−1.
Then
n
Y
i=1 i
X
j=1
wj si xj
!wi/sn
≥
n
X
j=1
wj sn
j
Y
i=1
xwii/sj.
Equality holds if and only ifx1 =x2 =· · ·=xn.
Remark 1. In terms of the sequenceti =s1+s2+· · ·+si, i= 1,2, . . . , n, it is not difficult to see that (2.1) is equivalent to the statement
t2i ≥ti−1ti+1, i= 2,3, . . . , n−1,
i.e., thatlogtiis concave, whereas (1.1) is equivalent to the assertion thatlogsi is concave. But we make no use of this alternative description of (2.1).
Before turning to the proof of Theorem2.1we show that (2.1) is implied by (1.1).
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page8of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
http://jipam.vu.edu.au
Lemma 2.2. Letw1, w2, . . . be a sequence of positive numbers, and define the sequences1, s2, . . . by
si =w1+w2+· · ·+wi, i= 1,2, . . . . Suppose
w1 s1 ≥ w2
s2 ≥ · · · ≥ wn
sn ≥ · · · . Then
s2k−wk+1
k−1
X
j=1
sj >0, k = 2,3, . . . .
Proof. The proof is by induction. To begin with, sincew2s2−w3s1 =w2s3− wss2 ≥0, we have that
s22−w3s1 =w1s2+w2s2−w3s1 ≥w1s2 >0.
So, suppose the claimed result holds for some m ≥ 2. Then, noting that, for i≥2,wisi−wi+1si−1 =wisi+1−wi+1si ≥0, we see that
s2m+1−wm+2
m
X
j=1
sj ≥ wm+2
wm+1sm+1sm−wm+2
m
X
j=1
sj
= wm+2
wm+1 sm+1sm−wm+1
m
X
j=1
sj
!
= wm+2
wm+1 s2m+wm+1sm−wm+1
m
X
j=1
sj
!
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page9of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
= wm+2
wm+1 s2m−wm+1
m−1
X
j=1
sj
!
>0,
by the induction assumption. The result follows.
We prove Theorem 2.1 by induction, and, to make productive use of the induction hypothesis, we need the following elementary result.
Lemma 2.3. LetA, B >0. Letp >1, q=p/(p−1). Then, for alls≥0, (A+Bs)p ≤(Aq+Bq)p−1(1 +sp),
with equality if and only ifs= (B/A)q−1.
Proof. The inequality is trivial if s = 0. Suppose s > 0. Exploiting the strict convexity oft→tq, it is clear that
A+Bs 1 +sp
q
=
A+ (Bs1−p)sp 1 +sp
q
≤ Aq+ (Bs1−p)qsp 1 +sp
= Aq+Bq 1 +sp ,
with equality if and only ifA =Bs1−p. The stated result follows quickly from this.
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page10of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
http://jipam.vu.edu.au
Corollary 2.4. Let p > 1, q = p/(p−1). LetA, B, C, D > 0. Then, for all t ≥0,
(A+Bt)p C+Dtp ≤ 1
CD(AqDq−1+BqCq−1)p−1, with equality if and only ift= (BC/AD)q−1.
We are now ready to deal with the proof of Theorem2.1.
For convenience, define the sequences of weighted averagesAk, Gkofx1, x2, . . . , xnby
Ak =
k
X
i=1
wi
sk xi, Gk =
k
Y
i=1
xwii/sk, k= 1,2, . . . , n.
We are required to prove that
n
X
i=1
wi
snGi ≤
n
Y
i=1
Awii/sn,
holds under condition (2.1), with equality if and only if x1 =x2 =· · ·=xn.
Proof. We prove this by induction. The result clearly holds for n = 1. More- over, as we mentioned in the introduction, a simple convexity argument estab- lishes that it also holds whenn = 2. We continue, therefore, with the assump- tion that n ≥ 3. Suppose the result holds for some positive integer m, with
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page11of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
1≤m≤n−1, so that, with X =
m
Y
i=1
Awii/sm,
then
m+1
X
i=1
wi sm+1
Gi = smPm i=1
wi
smGi+wm+1Gm+1
sm+1
≤ smX+wm+1Gm+1 sm+1
= (1−α)X+αY xαm+1, whereα=wm+1/sm+1and
Y =
m
Y
i=1
xwii/sm+1 =Gsmm/sm+1 =G1−αm .
In addition,
Am+1 = smAm+wm+1xm+1
sm+1 = (1−α)Am+αxm+1.
We claim now that
(1−α)X+αY xαm+1 ≤Xsm/sm+1Am+1wm+1/sm+1
=X1−α((1−α)Am+αxm+1)α,
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page12of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
http://jipam.vu.edu.au
i.e.,
((1−α)X+αY xαm+1)1/α
(1−α)Am+αxm+1 ≤X(1−α)/α.
By Corollary 2.4, withp = 1/α, A = (1−α)X, B = αY, C = (1−α)Am, D=α, q = 1/(1−α), the left-hand side does not exceed
(1−α)X1/(1−α)+αY1/(1−α)Aα/(1−α)m
(1−α)/α
Am ,
with equality if and only if
xm+1 =
Y Am
X
1/(1−α)
.
Thus, to finish the proof, we must establish that
(1−α)X1/(1−α)+αY1/(1−α)Aα/(1−α)m ≤XAα/(1−α)m ,
i.e., that
sm X
Am
α/(1−α)
+wm+1Y1/(1−α)
X ≤sm+1. In other words,
(2.2) sm Qm
i=1Awii/sm Am
!wm+1/sm
+wm+1
m
Y
i=1
xi Ai
wi/sm
≤sm+1,
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page13of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
with the additional assertion that there is equality if and only ifx1 =x2 =· · ·= xm. This inequality is of independent interest, and can be considered for its own sake. To prove it, consider the second term on the left-hand side of (2.2). This is equal to
wm+1Gm
X =wm+1sm v u u t
m
Y
i=1
xi Ai
wi
,
whence, by the convexity of the exponential function, bearing in mind thatsm = Pm
i=1wi, we see that this does not exceed wm+1
sm
m
X
i=1
wixi Ai .
Moreover, there is equality if and only if 1 = x1
A1
= xi Ai
, i= 1,2, . . . , m, i.e.,x1 =x2 =· · ·=xm.
Now we focus on the first term. To begin with, observe that X
Am = sm sQm
i=1Awii Asmm
= sm s
Qm−1 i=1 Awii Asmm−1
= sm v u u t
m−1
Y
i=1
Ai
Ai+1 si
.
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page14of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
http://jipam.vu.edu.au
Hence, once more by the convexity of the exponential function,
sm X
Am
α/(1−α)
=sm 1cm
m−1
Y
i=1
Ai Ai+1
si!wm+1/s2m
≤ wm+1
sm cm+
m−1
X
i=1
siAi Ai+1
!
= wm+1
sm cm+
m
X
i=2
si−1Ai−1
Ai
! ,
where
cm = s2m wm+1 −
m−1
X
i=1
si ≥0,
by hypothesis. Equality holds here if and only if 1 = Ai
Ai+1, i= 1,2, . . . , m−1,
i.e.,
si
i+1
X
j=1
wjxj =si+1
i
X
j=1
wjxj, i= 1,2, . . . , m−1, equivalently, if and only ifxm =· · ·=x2 =x1.
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page15of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
Combining our estimates we see that
sm X
Am
α/(1−α)
+wm+1Y1/(1−α) X
≤ wm+1
sm cm+
m
X
i=2
si−1Ai−1
Ai +
m
X
i=1
wixi
Ai
!
= wm+1
sm cm+w1 +
m
X
i=2
si−1Ai−1+wixi
Ai
!
= wm+1
sm cm+w1 +
m
X
i=2
siAi Ai
!
= wm+1
sm cm+
m−1
X
i=1
si+sm
!
= wm+1 sm
s2m
wm+1 +sm
=sm+1.
Thus (2.2) holds. Moreover, equality holds in (2.2) if and only if x1 = x2 =
· · ·=xm. Of course, (2.2) implies the inequality in Theorem2.1, by induction.
It therefore only remains to discuss the case of equality in this. But, if x1 = x2 =· · ·=xm, thenAm =X =x1, andY =xs1m/sm+1, whence equality holds throughout only if, in addition, xm+1 = Y1/(1−α = x1 also. But, clearly, the equality holds if all thex’s are equal. This finishes the proof.
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page16of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
http://jipam.vu.edu.au
References
[1] G. BENNETT, Factorizing the classical inequalities, Mem. Amer. Math.
Soc., 120 (1996), no. 576, viii+130 pp.
[2] G. BENNETT, An inequality for Hausdorff means, Houston J. Math., 25(4) (1996), 709–744.
[3] G. BENNETT, Summability matrices and random walk, Houston J. Math., 28(4) (2002), 865–898.
[4] T. CARLEMAN, Sur les functions quasi-analytiques, Fifth Scand. Math.
Congress (1923), 181–196.
[5] J.A. COCHRAN AND C.S. LEE, Inequalities related to Hardy’s and Heinig’s, Math. Proc. Camb. Phil. Soc., 96(1) (1984), 1–7.
[6] G.H. HARDY, Notes on some points of the integral calculus (LX), Mes- senger of Math., 54 (1925), 150–156.
[7] G.H. HARDY, J.E. LITTLEWOOD, AND G. PÓLYA, Inequalities, Cam- bridge University Press, 1934.
[8] H. HEINIG, Some extensions of Hardy’s inequality, SIAM J. Math. Anal., 6 (1975), 698–713.
[9] F. HOLLAND, On a mixed arithmetic-mean, geometric-mean inequality, Math. Competitions, 5 (1992), 60–64.
[10] F. HOLLAND, A strengthening of the Carleman-Hardy-Pólya inequality, Proc. Amer. Math. Soc., To appear.
An Inequality Between Compositions of Weighted
Arithmetic and Geometric Means
Finbarr Holland
Title Page Contents
JJ II
J I
Go Back Close
Quit Page17of17
J. Ineq. Pure and Appl. Math. 7(5) Art. 159, 2006
[11] M. JOHANSSON, L.E. PERSSON AND A. WEDESTIG, Carleman’s inequality—history, proofs and some new generalizations, J. Inequal. Pure
& Appl. Math., 4(3) (2003), Art. 53. [ONLINE:http://jipam.vu.
edu.au/article.php?sid=291].
[12] K. KEDLAYA, Proof of a mixed arithmetic-mean, geometric-mean in- equality, Amer. Math. Monthly, 101 (1994), 355–357.
[13] K. KEDLAYA, A Weighted Mixed-Mean Inequality, Amer. Math. Monthly, 106 (1999), 355–358.
[14] E.R. LOVE, Inequalities related to those of Hardy and of Cochran and Lee, Math. Proc. Camb. Phil. Soc., 99(1) (1986), 395–408.
[15] E.R. LOVE, Inequalities related to Carleman’s inequality, Inequalities, (Birmingham, 1987), 135–141, Lecture Notes in Pure and Appl. Math.
129, Dekker, New York, 1991.
[16] B. MONDANDJ.E. PE ˇCARI ´C, A mixed means inequality, Austral.Math.
Soc. Gaz., 23(2) (1996) 67–70.
[17] T.S. NANJUNDIAH, Sharpening of some classical inequalities, Math Stu- dent, 20 (1952), 24–25.
[18] B. OPIC AND P. GURKA, Weighted inequalities for geometric means, Proc. Amer. Math. Soc., 120 (1994), 771–779.