MonotonicBehaviourofRelativeIncrementsofPearsonDistributions

Similar documents
Hermite-Hadamard type inequalities for harmonically convex functions

On the Generalized Weighted Quasi-Arithmetic Integral Mean 1

Journal of Inequalities in Pure and Applied Mathematics

The asymptotic behavior of the real roots of Fibonacci-like polynomials

Euler, Ioachimescu and the trapezium rule. G.J.O. Jameson (Math. Gazette 96 (2012), )

Quadrature Rules for Evaluation of Hyper Singular Integrals

Research Article Fejér and Hermite-Hadamard Type Inequalities for Harmonically Convex Functions

The Modified Heinz s Inequality

The Hadamard s Inequality for s-convex Function

On some refinements of companions of Fejér s inequality via superquadratic functions

MAT137 Calculus! Lecture 20

Research Article Moment Inequalities and Complete Moment Convergence

Mathematics 1. (Integration)

The Shortest Confidence Interval for the Mean of a Normal Distribution

Research Article On Hermite-Hadamard Type Inequalities for Functions Whose Second Derivatives Absolute Values Are s-convex

Expected Value of Function of Uncertain Variables

ENGI 3424 Engineering Mathematics Five Tutorial Examples of Partial Fractions

A short introduction to local fractional complex analysis

Lecture 1. Functional series. Pointwise and uniform convergence.

7 - Continuous random variables

Continuous Random Variables

On a Method to Compute the Determinant of a 4 4 Matrix

New Expansion and Infinite Series

W. We shall do so one by one, starting with I 1, and we shall do it greedily, trying

MORE FUNCTION GRAPHING; OPTIMIZATION. (Last edited October 28, 2013 at 11:09pm.)

Hermite-Hadamard Type Inequalities for the Functions whose Second Derivatives in Absolute Value are Convex and Concave

Czechoslovak Mathematical Journal, 55 (130) (2005), , Abbotsford. 1. Introduction

Some estimates on the Hermite-Hadamard inequality through quasi-convex functions

8 Laplace s Method and Local Limit Theorems

FUNCTIONS OF α-slow INCREASE

The one-dimensional Henstock-Kurzweil integral

Co-ordinated s-convex Function in the First Sense with Some Hadamard-Type Inequalities

Proceedings of the International Conference on Theory and Applications of Mathematics and Informatics ICTAMI 2003, Alba Iulia

Realistic Method for Solving Fully Intuitionistic Fuzzy. Transportation Problems

Chapter 5 : Continuous Random Variables

Math 554 Integration

Reversals of Signal-Posterior Monotonicity for Any Bounded Prior

Chapter 3 Single Random Variables and Probability Distributions (Part 2)

Calculus in R. Chapter Di erentiation

A Generalized Inequality of Ostrowski Type for Twice Differentiable Bounded Mappings and Applications

The final exam will take place on Friday May 11th from 8am 11am in Evans room 60.

GlobalExistenceandUniquenessoftheWeakSolutioninKellerSegelModel

Journal of Inequalities in Pure and Applied Mathematics

1 Probability Density Functions

revised March 13, 2008 MONOTONIC SEQUENCES RELATED TO ZEROS OF BESSEL FUNCTIONS

Convex Sets and Functions

The Bochner Integral and the Weak Property (N)

ON A CONVEXITY PROPERTY. 1. Introduction Most general class of convex functions is defined by the inequality

Integral points on the rational curve

MATH362 Fundamentals of Mathematical Finance

20 MATHEMATICS POLYNOMIALS

GENERALIZATIONS OF WEIGHTED TRAPEZOIDAL INEQUALITY FOR MONOTONIC MAPPINGS AND ITS APPLICATIONS. (b a)3 [f(a) + f(b)] f x (a,b)

Math 118: Honours Calculus II Winter, 2005 List of Theorems. L(P, f) U(Q, f). f exists for each ǫ > 0 there exists a partition P of [a, b] such that

Journal of Inequalities in Pure and Applied Mathematics

Math 1431 Section 6.1. f x dx, find f. Question 22: If. a. 5 b. π c. π-5 d. 0 e. -5. Question 33: Choose the correct statement given that

ON SOME NEW INEQUALITIES OF HADAMARD TYPE INVOLVING h-convex FUNCTIONS. 1. Introduction. f(a) + f(b) f(x)dx b a. 2 a

Positive Solutions of Operator Equations on Half-Line

Arithmetic Mean Derivative Based Midpoint Rule

Lecture notes. Fundamental inequalities: techniques and applications

Appendix to Notes 8 (a)

Hermite-Hadamard-Fejér type inequalities for harmonically convex functions via fractional integrals

arxiv: v1 [math.ca] 28 Jan 2013

Solutions of Klein - Gordan equations, using Finite Fourier Sine Transform

Fact: All polynomial functions are continuous and differentiable everywhere.

Advanced Calculus: MATH 410 Notes on Integrals and Integrability Professor David Levermore 17 October 2004

S. S. Dragomir. 2, we have the inequality. b a

Partial Derivatives. Limits. For a single variable function f (x), the limit lim

Improper Integrals. MATH 211, Calculus II. J. Robert Buchanan. Spring Department of Mathematics

Exam 2, Mathematics 4701, Section ETY6 6:05 pm 7:40 pm, March 31, 2016, IH-1105 Instructor: Attila Máté 1

Research Article On The Hadamard s Inequality for Log-Convex Functions on the Coordinates

Geometrically Convex Function and Estimation of Remainder Terms in Taylor Series Expansion of some Functions

QUADRATURE is an old-fashioned word that refers to

On new Hermite-Hadamard-Fejer type inequalities for p-convex functions via fractional integrals

A Compound of Geeta Distribution with Generalized Beta Distribution

p-adic Egyptian Fractions

Section 6.1 INTRO to LAPLACE TRANSFORMS

Principles of Real Analysis I Fall VI. Riemann Integration

MATH20812: PRACTICAL STATISTICS I SEMESTER 2 NOTES ON RANDOM VARIABLES

BIFURCATIONS IN ONE-DIMENSIONAL DISCRETE SYSTEMS

Hamiltonian Cycle in Complete Multipartite Graphs

Journal of Inequalities in Pure and Applied Mathematics

A basic logarithmic inequality, and the logarithmic mean

Farey Fractions. Rickard Fernström. U.U.D.M. Project Report 2017:24. Department of Mathematics Uppsala University

A PROOF OF THE FUNDAMENTAL THEOREM OF CALCULUS USING HAUSDORFF MEASURES

Math 426: Probability Final Exam Practice

Homework 4. (1) If f R[a, b], show that f 3 R[a, b]. If f + (x) = max{f(x), 0}, is f + R[a, b]? Justify your answer.

DEFINITE INTEGRALS. f(x)dx exists. Note that, together with the definition of definite integrals, definitions (2) and (3) define b

The Riemann Integral

Frobenius numbers of generalized Fibonacci semigroups

S. S. Dragomir. 1. Introduction. In [1], Guessab and Schmeisser have proved among others, the following companion of Ostrowski s inequality:

WHEN IS A FUNCTION NOT FLAT? 1. Introduction. {e 1 0, x = 0. f(x) =

Houston Journal of Mathematics. c 1999 University of Houston Volume 25, No. 4, 1999

Homework Problem Set 1 Solutions

SUPERSTABILITY OF DIFFERENTIAL EQUATIONS WITH BOUNDARY CONDITIONS

FINALTERM EXAMINATION 2011 Calculus &. Analytical Geometry-I

Regulated functions and the regulated integral

Chapter 6. Riemann Integral

The Hadamard s inequality for quasi-convex functions via fractional integrals

Heavy tail and stable distributions

Calculus I-II Review Sheet

Transcription:

Globl Journl o Science Frontier Reserch: F Mthemtics nd Decision Sciences Volume 8 Issue 5 Version.0 Yer 208 Type : Double lind Peer Reviewed Interntionl Reserch Journl Publisher: Globl Journls Online ISSN: 2249-4626 & Print ISSN: 0975-5896 Monotonic ehviour o Reltive Increments o Person Distributions y University o otswn bstrct- Theory hs been developed in order to clssiy distributions c-cording to monotonic behviour o their reltive increment unctions. We pply the results to Person distributions. GJSFR-F Clssiiction: FOR Code: MSC 200: 2648 MonotonicehviouroReltiveIncrementsoPersonDistributions Strictly s per the complince nd regultions o: 208.. This is reserch/review pper, distributed under the terms o the Cretive Commons ttribution-noncommercil 3.0 Unported License http://cretivecommons.org/licenses/by-nc/3.0/), permitting ll non commercil use, distribution, nd reproduction in ny medium, provided the originl work is properly cited.

Re 2. Szbo, Z. : Investigtion o Reltive Increments o Distribution Functions. Publ. Mth. De-brecen 49 996), pp. 99-2 Monotonic ehviour o Reltive Increments o Person Distributions Let x) be the probbility density unction o continuous rndom vrible X with open support I. The corresponding distribution unction o X, Xx), is deined by F x) = x t)dt The reltive increment unction, h, o distribution unction, F, is deined by where c is positive constnt. bstrct- Theory hs been developed in order to clssiy distributions c-cording to monotonic behviour o their reltive increment unctions. We pply the results to Person distributions. I. Introduction hx) = F x + c) F x) F x) Lemm. Let F be twice dierentible distribution unction with F x) <, F x) = x) > 0 or ll x in I. We deine the unction Ψ s ollows Ψx) = F x) ). x) 2 x) I Ψ < Ψ > ), then it hs been proven tht the unction h strictly increses strictly decreses)[2] probbility distribution with probbility density unction x) is sid to be Person distribution i x) x) = Qx) qx) where Qx) = x + nd qx) = x 2 + bx + c nd,,, b, c re rel constnts with 2 + b 2 + c 2 > 0, 2 + 2 > 0 [] Globl Journl o Science Frontier Reserch Volume XVIII Issue V V ersion I Yer 208 F ) The ollowing our theorems hve been proven beore nd we will use them to ormulte results bout Person distributions. uthor: Deprtment o Mthemtics, University o otswn. e-mil: reikeletsengsn@mopipi.ub.bw 208 Globl Journls

Monotonic ehviour o Reltive Increments o Person Distributions II. Theorem. I the probbility density unction hs the ollowing properties..) I = r, s) R is the lrgest inite or ininite open intervl in which > 0..2) There exists m in I t which is continuous nd m) = 0.3) > 0 in r, m) nd < 0 in m, s).4) is twice dierentible in m, s) ) ).5) = d dx > 0 in m, s), Then the corresponding continuous reltive increment unction, h, behves s ollows: I Ψs ) = lim Ψx) exists, then x s h strictly increses in I i Ψs ) h strictly increses in r, y) nd strictly decreses in y, s) or some y in I i Ψs ) >.[2] Re Globl Journl o Science Frontier Reserch Volume XVIII Issue V V ersion I Yer 208 2F ) III. Theorem 2. I the probbility density unction hs the ollowing properties. 2.) I = r, s) R is the lrgest inite or ininite open intervl in which > 0. 2.2) m = r 2.3) < 0 in r, s) 2.4) is twice dierentible in r, s) 2.5) ) = d dx ) < 0 in r, s), then r is inite. i) I Ψr + ) < or Ψr + ) = nd Ψ < in some right neighbourhood o r), then Ψ < in I nd the corresponding reltive increment unction strictly increses in I. ii) I Ψr + ) >, then I Ψs ), then Ψ > nd the reltive increment unction strictly decreses in I. I Ψs ) <, then Ψ > in r, y) nd Ψ < in y, s) or some y I, so the reltive increment unction strictly decreses irst nd then strictly increses.[2] Monotonic ehviour o Reltive Increments o Person Distributions Theorem 3. I the probbility density unction hs the ollowing properties. 3.) I = r, s) R is the lrgest inite or ininite open intervl in which > 0. 3.2) m = r 3.3) < 0 in r, s) 3.4) is twice dierentible in r, s) 3.5) ) = d dx ) > 0 in r, s), then r is inite i) I Ψm + ) > or Ψm + ) = nd Ψ > in some right neighbourhood o m), then Ψ > in I nd the corresponding reltive increment unction strictly decreses in I. ii) I Ψm + ) <, then I Ψs ) <, then Ψ < nd the reltive increment unction strictly increses in I. I Ψs ) >, then Ψ < in m, x) nd Ψ > in x, s) or some x I, so the reltive increment unction strictly increses irst nd then strictly decreses.[4] Theorem 4. Let x) be probbility density unction o U type distribution where.).4) re ulilled. Suppose ) > 0 in m, y) nd ) < 0 in y, s) or some y m, s). Then the corresponding continuous reltive increment unction, h, behves s ollows: I Ψy) < nd Ψs ) < then strictly increses in I or 2. Szbo, Z. : Investigtion o Reltive Increments o Distribution Functions. Publ. Mth. De-brecen 49 996), pp. 99-2 208 Globl Journls

Monotonic ehviour o Reltive Increments o Person Distributions Re 4. Szbo Z. : On Monotonic ehviour o Reltive Increments o Unimodl Distributions. I Ψy) then i Ψs ) > h strictly increses in r, x 0 ) or some x 0 m, y) nd strictly decreses in x 0, s). i Ψs ) < then h strictly increses in r, x ) or some x m, y) nd strictly decreses in x, x 2 ) or some x 2 y, s) then inlly strictly increses in x 2, s).[4] Theorem 5. Let x) be probbility density unction o J type distribution. ssume m = r,.),.3),.4) re ulilled. Suppose ) < 0 in m, Y ) nd ) > 0 in Y, s) or some Y m, s). Then the corresponding continuous reltive increment unction, h, behves s ollows: i Ψm + ) = lim Ψx) exists, then x m + I Ψm + ) < nd Ψs ) < then h strictly increses in I or Ψs ) > then h strictly increses in m, x 3 ) or some x 3 y, s) nd strictly decreses in x 3, s) I Ψm + ) nd Ψy) > then h strictly decreses in I Ψy) < then i i Ψs ) < then h strictly decreses in m, x 4 ) or some x 4 m, y) nd strictly increses in x 4, s) IV. i Ψs ) > then h strictly decreses in m, x 5 ) or some x 5 m, y) nd strictly increses in x 5, x 6 ) or some x 6 y, s) then inlly strictly decreses in x 6, s) [4] V. MIN RESULTS For the next two theorems, we use the nottion used in [3] Theorem 6. Let x) be density unction o Person distribution where.).4) re ulilled Let M = b..c, L =. 2.M, D =.L, D = D nd b = b 2 or 0. ) I = 0 nd < 0 then s + b 0. 2) I. > 0 nd q ) 0 nd D 0 then Y m) 0 nd 0 Y s) 2D 3) I. < 0 nd q ) 0 nd D 0 then 0 Y m) 2D nd Y s) 0 where Y v) = v + ) + D. Then ll the ssertions o Theorem 4 hold. Proo. y theorem 4, it is suicient to prove tht or some y in m, s) We hve where nd ) > 0 in m, y) nd ) < 0 in y, s) ) = q Q ) = px) Q 2 px) = x 2 + 2x + M ) > 0 i px) > 0 nd ) < 0 i px) < 0 I 0 nd D 0 then the roots o px) re Globl Journl o Science Frontier Reserch Volume XVIII Issue V V ersion I Yer 208 3F ) x = D) ) nd x 2 = +D) ) where D = D 2. 208 Globl Journls

Monotonic ehviour o Reltive Increments o Person Distributions Cse : 0, 0 nd q ) 0, so q nd Q hve no common zero... I. > 0, then px) is convex. I D 0 then there will be two rel roots x nd x 2 where x x 2. i nd only i px) > 0 in m, y) nd px) < 0 in y, s) which mens tht m x, x = y nd x < s < x 2 Y m) < 0 nd 0 Y s) 2D Globl Journl o Science Frontier Reserch Volume XVIII Issue V V ersion I Yer 208 4F ) MONOTONIC EHVIOUR OF RELTIVE INCREMENTS OF PERSON DISTRIUTIONS 5.2. I. < 0, then px) is concve. i nd only i which mens tht Cse 2 = 0 nd 0 thereore nd I > 0 then q Q ) > 0 i x > I < 0 then q Q ) > 0 i x < px) > 0 in m, y) nd px) < 0 in y, s) x m x 2 nd s x 2 = y 0 Y m) 2D nd Y s) 0 q Q = x2 + bx + c q Q ) = 2x + b nd q Q ) < 0 i x <. I < 0 then s > 2 nd q Q ) < 0 i x >. I > 0 then s > 2 Cse 3 I = 0, then px) = b. I b > 0 then Theorem pplies. I b < 0 then there is no cse like it s seen in remrk 2. in [3]. Theorem 7 Let x) be the density unction o Person distribution with m = r,.),.3),.4) re ulilled nd M, L, D deined s in theorem 5. ) I = 0 nd < 0 then m + b 0. 2) I. > 0 nd q ) 0 nd D 0 nd Y m) 0 nd 0 Y s) 2D 3) I. < 0 nd q ) 0 nd D 0 nd 0 Y m) 2D nd Y s) 0 where Y v) = v + ) D. Then ll the ssertions o Theorem 5 hold. Proo. y theorem 5, it is suicient to prove tht or some y in m, s). Cse : ) < 0 in m, y) nd ) > 0 in y, s) 0 nd q ) 0, so q nd Q hve no common zero... I. > 0, then px) is convex. I D 0 then there will be two rel roots x nd x 2 where x < x 2. i nd only i px) < 0 m, y) nd px) > 0 y, s) m = x, x 2 = y nd s > x 2 Re 3. Szbo, Z. : Reltive Increments o Person Distributions. ct Mth. c. Ped. Nyiregyh. 5 999), pp. 45-54. [Electronic Journl, websites www.bgyt.hu/mpn, www.emis.de/journls] 208 Globl Journls

Monotonic ehviour o Reltive Increments o Person Distributions which mens tht Y s) > 0 VI. Monotonic ehviour o Reltive Increments o Person Distributions Re 3. Szbo, Z. : Reltive Increments o Person Distributions. ct Mth. c. Ped. Nyiregyh. 5 999), pp. 45-54. [Electronic Journl, websites www.bgyt.hu/mpn, www.emis.de/journls].2. I. < 0, then px) is concve. i nd only i Cse 2 = 0 thereore nd px) < 0 m, y) nd px) > 0 y, s) m < x which mens tht Y m) > 0, y = x nd s = x 2 I > 0 then q Q ) < 0 i x < I < 0 then q Q ) < 0 i x > q Q = x2 + bx + c q Q ) = 2x + b nd q Q ) > 0 i x >. I > 0 then m < 2 nd q Q ) > 0 i x <. I < 0 then m < 2 Cse 3 I = 0, then px) = b. I b < 0 then Theorem 2 pplies. > 0 then Theorem 3 pplies. I b We stte the ollowing Lemm s it helps in clssiying distributions. These re outlined nd proven in [3]. Lemm 2. Let x) be the density unction o Person distribution, then then: I Let s = nd = lim x x.x) I. I = = = 0 nd b + 0 then Ψ ) = b+) I.2 I 0 nd = 0 then Ψ ) = I.3 I.. + ) 0 nd = 0 then Ψ ) = +) I.4 I = = 0, b. = 0) or = 0, 0) or.. 0) then Ψ ) = 0 II Let s be inite number nd let lim x s x) = 0 II. I [.Qs).qs) 0] or [.q s) 0, Qs) = qs) = 0] or [ = 0, qs) 0] or [.Qs) 0, qs) = q s) = 0], then Ψs ) = Exmple We hve II.2 I 0, q ).Qs).[Qs) + q s)] 0, qs) = 0 then Ψs ) = Qs) [Qs)+q s)] II.3 I 0, q ).qs) 0, Qs) = 0 then Ψs = 0) II.4 I 0, q ) = qs) = 0, + 0 nd either Qs) = q s) = 0 or Qs) 0 then Ψs ) = +) II.5 I = = qs) = 0, q s).b + ) 0 then Ψs ) = VII. x) = Γn ) x n exp k x ), n = 2, 3, 4,..., I = 0, ) = nx+k x 2 so = n, = k, =, b = 0, c = 0 b+ Globl Journl o Science Frontier Reserch Volume XVIII Issue V V ersion I Yer 208 5F ) 208 Globl Journls

Globl Journl o Science Frontier Reserch ) Volume XVIII Issue V V ersion I 6F Yer 208 Monotonic ehviour o Reltive Increments o Person Distributions Theorem 6 prt 3 pplies since q ) = k2 n 0 nd. = n < 0. 2 mode = k n < y = 2k n, so it s U-type distribution with px) = x2k nx) > 0 or x < 2k n nd px) < 0 or x > 2k n lim xx) = lim x x Γn ) x n+) exp k/x) = 0 so I.3 in Lemm 2 pplies giving ψ ) = + = n n+ > This mens tht ψy) > since i it ws less thn then ccording to theory ψx) would decrese nd remin less thn in y, s). So the reltive increment unction increses nd then decreses. Exmple 2 x) = K + x 2 ) n exp rctnx)), n > 2 We hve, K is constnt nd I =, ) = 2nx x 2 + so = 2n, =, =, b = 0, c = Theorem 6 prt 3 pplies since q ) = 0 nd. = 2n < 0. mode = 2n < y = + 4n 2 + 2n, so it s U-type distribution with px) = 2 nx 2 + x n) > 0 or x < + 4n 2 + 2n nd px) < 0 or x > + 4n 2 + 2n lim xx) = 0 x so I.3 in Lemm 2 pplies giving ψ ) = + = 2n 2n+ > This mens tht ψy) > since i it ws less thn then ccording to theory ψx) would decrese nd remin less thn in y, s). So the reltive increment unction increses nd then decreses. VIII. Reerences Réérences Reerencis. Johnson N. nd Kotz S. : Distributions in Sttistics. Continuous Univrite Distributions, vol.,2, Houghton Mi_in, oston - New York, 970. 2. Szbo, Z. : Investigtion o Reltive Increments o Distribution Functions. Publ. Mth. De-brecen 49 996), pp. 99-2 3. Szbo, Z. : Reltive Increments o Person Distributions. ct Mth. c. Ped. Nyiregyh. 5 999), pp. 45-54. [Electronic Journl, websites www.bgyt.hu/mpn, www.emis.de/journls] 4. Szbo Z. : On Monotonic ehviour o Reltive Increments o Unimodl Distributions. 5. ct Mth. cd. Ped. Nyiregyh. 22 2006), pp. 3-9. [Electronic Journl, website www.emis.de/journls] Notes 208 Globl Journls