Kragujevac Journal of Mathematics Volume 36 Number 1 (2012), Pages 73–76. SOME REMARKS ON RESULTS OF MORTICI TSERENDORJ BATBOLD Abstract. In this paper, we establish new generalizations of Turán-type inequality involving the Gamma and Polygamma functions. 1. Introduction P. Turán ([1]) proved that the Legendre polynomials Pn (x) satisfy the determinantal inequality (1.1) ¯ ¯ P (x) Pn+1 (x) ¯ n ¯ ¯ Pn+1 (x) Pn+2 (x) ¯ ¯ ¯ ¯ ≤ 0, ¯ −1 ≤ x ≤ 1 where n = 0, 1, 2, . . . and equality occurs only if x = ±1. This classical result has been extended in several directions: ultraspherical polynomials, Lagguerre and Hermite polynomials, Bessel functions of the first kind, modified Bessel functions, etc. Today there is a huge literature on Turán inequalities, since they have important applications in complex analysis, number theory, combinatorics, theory of mean-values, or statistics and control theory. Recently, Mortici ([2], [3]) proved some Turán-type inequalities for some special functions as well as the polygamma functions. The aim of this paper is to prove new generalizations of Turán-type inequalities for the Gamma and Polygamma functions in ([2],[3]). 2. Main Results P = 1 and Lemma 2.1 (Hölder’s inequality). If p1 , . . . , pn > 0 are such that ni=1 p−1 i f1 , . . . , fn are non-negative functions such that these integrals exist, then the following Key words and phrases. Turán-type inequalities, Gamma function, Polygamma functions, Hölder’s inequality. 2010 Mathematics Subject Classification. 26D07, 33B15. Received: October 16, 2011 . 73 74 TSERENDORJ BATBOLD inequality holds: n µZ ∞ Y pi (2.1) pi fi (t)dt 0 i=1 Z ∞ ÃY n ¶1 ≥ 0 ! fi (t) dt. i=1 In what follows, we use the integral representations, for x > 0 and n = 1, 2, . . . (n) (2.2) Γ (2.3) (n) ψ (x) = 0 e−t tx−1 logn tdt, n+1 (x) = (−1) Z ∞ n −tx t e 1 − e−t 0 1 Z ∞ tx−1 dt, ζ(x) = Γ(x) 0 et − 1 (2.4) and (2.5) Z ∞ θ(x) = (−1) n Z ∞µ 0 dt, x > 1, ¶ 1 1 1 n−1 −tx − + t e dt, t e −1 t 2 where Γ is the gamma function, ψ (n) is the n-th polygamma function and ζ is the Reimann-zeta function. We first give a generalization of Mortici ([2, Theorem 2.1], [3, Theorem 2.2]). Theorem 2.1. Let p1 , . . . , pn be conjugate parameters such that pi > 1, i = 1, . . . , n, P P and ni=1 p1i = 1, and let m1 , . . . , mn ≥ 1 be integers such that ni=1 mpii is an integer. Then the following inequality holds for xi > 0, i = 1, . . . , n: ¯ ³P n ¯ n ¯ ¯1 Y ¯ (mi ) ¯ ¯ pi ¯ψ (xi )¯ ≥ ¯ψ i=1 ¯ (2.6) i=1 mi pi ´Ã !¯ n X xi ¯¯ ¯. pi ¯ i=1 Proof. By Hölder’s inequality (2.1) and (2.3), we have n ¯ ¯1 Y ¯ (mi ) ¯ψ (xi )¯¯ pi = i=1 n Y ÃZ ∞ tmi e−txi i=1 ≥ Z ∞ 1−e 0 0 n Y i=1 Pn Z ∞ t t mi pi mi i=1 pi dt −t e (1 − !1 −txi pi 1 e−t ) pi −t pi ³P dt n xi i=1 pi ´ e dt −t 1 − e 0 ´Ã ¯ ³P !¯ n n mi ¯ X xi ¯¯ ¯ i=1 pi = ¯¯ψ ¯. ¯ i=1 pi = Hence (2.6) is valid. ¤ The next result is a generalization of Mortici ([2, Theorem 2.2], [3, Theorem 2.1, 2.3]). SOME REMARKS ON RESULTS OF MORTICI 75 Theorem 2.2. Let p1 , . . . , pn be conjugate parameters such that pi > 1, i = 1, . . . , n, P P and ni=1 p1i = 1, and let m1 , . . . , mn ≥ 1 be integers such that ni=1 mpii is an integer. Then the following inequalities hold: n ³ Y (2.7) Γ (mi ) ³P ´1 pi (xi ) ≥Γ n mi i=1 pi ´Ã xi > 0, i = 1, . . . , n, (2.8) , xi > 1, i = 1, . . . , n, !¯ n X xi ¯¯ ¯, pi ¯ xi > 0, i = 1, . . . , n. i=1 (ζ(xi )) 1 pi Γ ³P n xi i=1 pi ≥Q n 1 pi i=1 (Γ(xi )) i=1 ¯ ³P n ¯ n ¯1 ¯ Y ¯ (mi ) ¯ pi ¯ ¯θ (xi )¯ ≥ ¯θ i=1 ¯ (2.9) ´ mi pi ζ ! , i=1 n Y n X xi pi à n ! X xi i=1 ´Ã pi i=1 i=1 Proof. We just use Hölder’s inequality (2.1) and the proofs are similar to the proof of Theorem 2.1. We omit the details. ¤ The next result is a generalization of Mortici ([2, Theorem 3.1]). Theorem 2.3. Let p1 , . . . , pn be conjugate parameters such that pi > 1, i = 1, . . . , n, P P and ni=1 p1i = 1, and let m1 , . . . , mn ≥ 0 be even integers such that nj=1 mpjj is even integers. Then the following inequality holds for xi > 0, i = 1, . . . , n: ³P (2.10) n mi i=1 pi exp Γ ´Ã ! n X xi pi i=1 n ³ Y ´1 exp Γ(mi ) (xi ) ≤ pi . i=1 Proof. Using (2.2) and Weighted AM-GM inequality n X Γ(mi ) (xi ) pi i=1 = n X i=1 = ≥ ³P −Γ n mi i=1 pi i=1 pi 0 i=1 pi Z ∞ ÃY n 0 ! n X xi pi ³P Z ∞ n 1 Z ∞ −t xi −1 mi −t e t log tdt − e t i=1 ³P Z ∞ ÃX n 1 xi − nj=1 0 ´Ã t t xi − p1 pi i ³P ´ xj pj xj n j=1 pj 0 ³P log ´ log mi − mi − p1 pi i mj n j=1 pj xi pi ´ ´ −1 ³P log ! t − 1 e−t t ³P mj n j=1 pj ´ n mi i=1 pi ³P n xi i=1 pi ³P ! t − 1 e−t t ´ tdt ´ −1 n xi i=1 pi ³P log ´ −1 n mi i=1 pi ³P log ´ n mi i=1 pi tdt ´ tdt i=1 = 0. The conclusion follows by exponentiating the inequality n X Γ(mi ) (xi ) i=1 pi ³P n ≥Γ mi i=1 pi ´Ã n X xi i=1 pi ! . ¤ 76 TSERENDORJ BATBOLD Acknowledgement: The author wish to thank the anonymous referee for through review and insightful suggestions. References [1] P. Turán, On the zeros of the polynomials of Legendre, Casopis pro Pestovani Mat. a Fys. 75 (1950), 113–122. [2] C. Mortici, Turn-type inequalities for the gamma and polygamma functions, Acta Univ. Apulensis. 23 (2010), 117–121. [3] C. Mortici, New inequalities for some special functions via the Cauchy-Buniakovsky-Schwarz inequality, Tamkang J. Math. 42 (2011), no. 1, 53–57. Institute of Mathematics National University of Mongolia P.O. Box 46A/104, Ulaanbaatar 14201, Mongolia E-mail address: [email protected]
© Copyright 2026 Paperzz