A NEW GENERALISATION OF SAM-SOLAI S MULTIVARIATE ADDITIVE GAMMA DISTRIBUTION*

Similar documents
Research Journal of Pure Algebra -2(12), 2012, Page: Available online through ISSN

Folding of Regular CW-Complexes

A Note on Estimability in Linear Models

On Properties of the difference between two modified C p statistics in the nested multivariate linear regression models

Grand Canonical Ensemble

Analyzing Frequencies

COMPLEX NUMBER PAIRWISE COMPARISON AND COMPLEX NUMBER AHP

The Hyperelastic material is examined in this section.

ON THE COMPLEXITY OF K-STEP AND K-HOP DOMINATING SETS IN GRAPHS

ST 524 NCSU - Fall 2008 One way Analysis of variance Variances not homogeneous

Review - Probabilistic Classification

10/7/14. Mixture Models. Comp 135 Introduction to Machine Learning and Data Mining. Maximum likelihood estimation. Mixture of Normals in 1D

Chapter 6 Student Lecture Notes 6-1

te Finance (4th Edition), July 2017.

Lucas Test is based on Euler s theorem which states that if n is any integer and a is coprime to n, then a φ(n) 1modn.

PARTIAL DISTRIBUTION FUNCTION AND RADIAL STATISTICAL COEFFICIENTS FOR Be-ATOM

UNIT 8 TWO-WAY ANOVA WITH m OBSERVATIONS PER CELL

Journal of Theoretical and Applied Information Technology 10 th January Vol. 47 No JATIT & LLS. All rights reserved.

Stress-Based Finite Element Methods for Dynamics Analysis of Euler-Bernoulli Beams with Various Boundary Conditions

The Equitable Dominating Graph

arxiv: v1 [math.pr] 28 Jan 2019

VISUALIZATION OF DIFFERENTIAL GEOMETRY UDC 514.7(045) : : Eberhard Malkowsky 1, Vesna Veličković 2

HORIZONTAL IMPEDANCE FUNCTION OF SINGLE PILE IN SOIL LAYER WITH VARIABLE PROPERTIES

NEW APPLICATIONS OF THE ABEL-LIOUVILLE FORMULA

Decision-making with Distance-based Operators in Fuzzy Logic Control

ANALYTICITY THEOREM FOR FRACTIONAL LAPLACE TRANSFORM

PROOF OF FIRST STANDARD FORM OF NONELEMENTARY FUNCTIONS

4.8 Huffman Codes. Wordle. Encoding Text. Encoding Text. Prefix Codes. Encoding Text

2. Grundlegende Verfahren zur Übertragung digitaler Signale (Zusammenfassung) Informationstechnik Universität Ulm

Economics 600: August, 2007 Dynamic Part: Problem Set 5. Problems on Differential Equations and Continuous Time Optimization

SCITECH Volume 5, Issue 1 RESEARCH ORGANISATION November 17, 2015

Reliability of time dependent stress-strength system for various distributions

Independent Domination in Line Graphs

Soft k-means Clustering. Comp 135 Machine Learning Computer Science Tufts University. Mixture Models. Mixture of Normals in 1D

External Equivalent. EE 521 Analysis of Power Systems. Chen-Ching Liu, Boeing Distinguished Professor Washington State University

September 27, Introduction to Ordinary Differential Equations. ME 501A Seminar in Engineering Analysis Page 1. Outline

Matched Quick Switching Variable Sampling System with Quick Switching Attribute Sampling System

Level crossing rate of relay system over LOS multipath fading channel with one and two clusters in the presence of co-channel interference

CHAPTER 7d. DIFFERENTIATION AND INTEGRATION

Heisenberg Model. Sayed Mohammad Mahdi Sadrnezhaad. Supervisor: Prof. Abdollah Langari

LOGIT ANALYSIS. A.K. VASISHT Indian Agricultural Statistics Research Institute, Library Avenue, New Delhi

From Structural Analysis to FEM. Dhiman Basu

Group Codes Define Over Dihedral Groups of Small Order

CHAPTER 33: PARTICLE PHYSICS

8-node quadrilateral element. Numerical integration

Outlier-tolerant parameter estimation

A NON-LINEAR MODEL FOR STUDYING THE MOTION OF A HUMAN BODY. Piteşti, , Romania 2 Department of Automotive, University of Piteşti

Case Study of Cascade Reliability with weibull Distribution

Dr. Shalabh Department of Mathematics and Statistics Indian Institute of Technology Kanpur

FREE VIBRATION ANALYSIS OF FUNCTIONALLY GRADED BEAMS

SCHUR S THEOREM REU SUMMER 2005

(Upside-Down o Direct Rotation) β - Numbers

Here is the rationale: If X and y have a strong positive relationship to one another, then ( x x) will tend to be positive when ( y y)

Lecture 14. Relic neutrinos Temperature at neutrino decoupling and today Effective degeneracy factor Neutrino mass limits Saha equation

Logistic Regression I. HRP 261 2/10/ am

Ερωτήσεις και ασκησεις Κεφ. 10 (για μόρια) ΠΑΡΑΔΟΣΗ 29/11/2016. (d)

MP IN BLOCK QUASI-INCOHERENT DICTIONARIES

Strategies evaluation on the attempts to gain access to a service system (the second problem of an impatient customer)

Lecture 3: Phasor notation, Transfer Functions. Context

4D SIMPLICIAL QUANTUM GRAVITY

Optimal Ordering Policy in a Two-Level Supply Chain with Budget Constraint

A Probabilistic Characterization of Simulation Model Uncertainties

The Fourier Transform

Discrete Shells Simulation

Basic Electrical Engineering for Welding [ ] --- Introduction ---

Special Random Variables: Part 1

Naresuan University Journal: Science and Technology 2018; (26)1

Relate p and T at equilibrium between two phases. An open system where a new phase may form or a new component can be added

3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X

Comparing two Quantiles: the Burr Type X and Weibull Cases

EEC 686/785 Modeling & Performance Evaluation of Computer Systems. Lecture 12

Green Functions, the Generating Functional and Propagators in the Canonical Quantization Approach

j) = 1 (note sigma notation) ii. Continuous random variable (e.g. Normal distribution) 1. density function: f ( x) 0 and f ( x) dx = 1

IV. Transport Phenomena Lecture 35: Porous Electrodes (I. Supercapacitors)

The graph of y = x (or y = ) consists of two branches, As x 0, y + ; as x 0, y +. x = 0 is the

3.4 Properties of the Stress Tensor

10. The Discrete-Time Fourier Transform (DTFT)

:2;$-$(01*%<*=,-./-*=0;"%/;"-*

26 The University of Texas at Austin

ON THE INTEGRAL INVARIANTS OF KINEMATICALLY GENERATED RULED SURFACES *

Estimation: Part 2. Chapter GREG estimation

Hardy-Littlewood Conjecture and Exceptional real Zero. JinHua Fei. ChangLing Company of Electronic Technology Baoji Shannxi P.R.

First Year Examination Department of Statistics, University of Florida

Available online at ScienceDirect. Procedia Economics and Finance 17 ( 2014 ) 39 46

The root mean square value of the distribution is the square root of the mean square value of x: 1/2. Xrms

1) They represent a continuum of energies (there is no energy quantization). where all values of p are allowed so there is a continuum of energies.

ANALYSIS IN THE FREQUENCY DOMAIN

Fitting bivariate losses with phase-type distributions

On Parameter Estimation of the Envelope Gaussian Mixture Model

Supplementary material: Margin based PU Learning. Matrix Concentration Inequalities

SPECTRUM ESTIMATION (2)

Chat eld, C. and A.J.Collins, Introduction to multivariate analysis. Chapman & Hall, 1980

1973 AP Calculus AB: Section I

1- Summary of Kinetic Theory of Gases

LINEAR REGRESSION ANALYSIS. MODULE VIII Lecture Indicator Variables

MODEL QUESTION. Statistics (Theory) (New Syllabus) dx OR, If M is the mode of a discrete probability distribution with mass function f

ON EISENSTEIN-DUMAS AND GENERALIZED SCHÖNEMANN POLYNOMIALS

On Selection of Best Sensitive Logistic Estimator in the Presence of Collinearity

Statistics and Probability Theory in Civil, Surveying and Environmental Engineering

Α complete processing methodology for 3D monitoring using GNSS receivers

Transcription:

A NEW GENERALISATION OF SAM-SOLAI S MULTIVARIATE ADDITIVE GAMMA DISTRIBUTION* Dr. G.S. Davd Sam Jayakumar, Assstant Profssor, Jamal Insttut of Managmnt, Jamal Mohamd Collg, Truchraall 620 020, South Inda, Inda. E-mal: samjaya77@gmal.com Dr. A. Solaraju, Assocat Profssor, Dt. of Mathmatcs, Jamal Mohamd Collg, Truchraall 620 020, South Inda, Inda. E-mal:Solarama@yahoo.co.n A.Sulthan, Rsarch scholar, Jamal Insttut of Managmnt, Truchraall 620 020 South Inda, Inda, E-mal: Sulthan90@gmal.com Abstract: Ths ar roosd a nw gnralzaton of boundd Contnuous multvarat symmtrc robablty dstrbutons. In ths ar, w vsualz a nw gnralzaton of Sam-Sola s Multvarat addtv Gamma dstrbuton from th un-varat two aramtrs Gamma dstrbuton. Furthr, w fnd ts Margnal, Multvarat Condtonal dstrbutons, Multvarat Gnratng functons, Multvarat survval, hazard functons and also dscussd t s scal cass. Th scal cass ncluds th transformaton of Sam-Sola s Multvarat addtv Gamma dstrbuton nto Multvarat Ch-squar dstrbuton, Multvarat Erlang dstrbuton, Two aramtr Multvarat Gamma dstrbuton, Multvarat Invrs Gamma dstrbuton, Multvarat log Gamma dstrbuton and Multvarat Nagakam-m dstrbuton. Morovr, t s found that th bvarat corrlaton btwn two Gamma random varabls urly dnds on th sha aramtr and w smulatd and stablshd slctd standard bvarat gamma corrlaton bounds from 900 dffrnt combnatons of valus for sha aramtr. Kywords: Sam-Sola s Multvarat Gamma dstrbuton, Transformaton, Multvarat Ch-squar dstrbuton, Multvarat Erlang dstrbuton, Two aramtr Multvarat Gamma dstrbuton, Multvarat Invrs Gamma dstrbuton, Multvarat log Gamma dstrbuton, Multvarat Nagakam-m dstrbuton, corrlaton bounds Introducton: Chryan (94) ntroducd a b-varat Gamma ty dstrbuton functon wth assumton of th Gamma random varabls ar corrlatd and smlarly Ramabhadran [95] roosd a multvarat Gamma ty dstrbuton n th xonntal famly of functons. Morovr Krshnamoorthy t al. (95) contnud th work of chryan, Ramabhadran and roosd a smlar ty of multvarat Gamma dstrbuton. On th othr hand, Sarmanov (968) roosd a gnralzd Gamma dstrbuton wth th assumton of symmtrcty among random varabls and Gavr (970) stablshd th mxtur of multvarat Gamma dstrbuton. Johnson t al (972, 2000) hghlghtd th Multvarat systm of Gamma dstrbuton and Dussauchoy t al (975) ntroducd a Multvarat Gamma dstrbuton whos margnal ar also followd a unvarat Gamma laws. Bckr t al(98) studd th xtnson of gamma dstrbuton for th bvarat cas and smlarly D Est(98) also dscrbd th Morgnstrn ty Gnralzaton of bvarat Gamma dstrbuton.kowalczyk t al(989) conductd a n-dth study about th rorts of Multvarat Gamma dstrbuton namly thr sha, stmaton of aramtrs and Matha(99,992) studd a dffrnt form of multvarat Gamma dstrbuton. Basd on th ast and rsnt ltraturs, th authors roosd a nw gnralzaton of boundd Contnuous multvarat symmtrc robablty dstrbutons wth scal rfrnc to th Gamma law and t s dscussd n th nxt scton. Thus th logcal gnralzaton of unvarat robablty dstrbuton for a Multvarat cas s an ntrstng task on th art of statstcans. Th gnralzaton of unvarat two aramtr Gamma dstrbuton to ts Multvarat cas basd on th addtv ty dstrbuton s dscussd. *Mathmatcs Subjct Classfcaton. Prmary 62H0; Scondary 62E5 Global Publshng Comany 39

G. S. DAVID SAM JAYAKUMAR, A. SOLAIRAJU, A. SULTHAN Scton : Sam-Sola s Multvarat Addtv Gamma dstrbuton Dfnton.: Lt X, X 2, X3, X b th random varabls followd Contnuous unvarat Gamma dstrbuton wth sha aramtr k and scal aramtr for all (= to ). Thn th dnsty functon of th Multvarat Sam Sola s addtv Gamma dstrbuton s dfnd as x f ( x, x, x, x ) {( ) ( )} k k x -- () whr 0, k 0. x Thorm.2: Th cumulatv dstrbuton functon of th Sam-Sola s Multvarat addtv Gamma dstrbuton s dfnd by x x2 x x 3 k u ( u ) 0 0 0 0 k F( x, x, x, x ) {( ) ( )} du du du du -- (2) Whr 0 u x, k 0 x k k u x u x k 0 F( x, x, x, x ) ( ){ ( du ) ( ))} (,, ) x x k F( x, x2, x3, x ) ( ){ ( ) ( )} x whr x k k u u ( x,, k ) du k 0 s th lowr ncomlt Gamma ntgral of th random varabl. Thorm.3: Th Probablty dnsty functon of Sam-Sola s Multvarat addtv Condtonal Gamma dstrbuton of X on X 2, X 3, X s P ( x ) {( ) ( )} k k k ( x) f ( x / x, x, x ) ( ) ( 2) k 2 x -- (3) whr 0 x, k 0 Proof: It s obtand from f ( x / x, x, x ) f ( x, x, x, x ) f ( x, x, x ) 2 3 Global Publshng Comany 40

A NEW GENERALIZATION Thorm.4: Man and Varanc of Sam - Sola s Multvarat addtv Condtonal Gamma dstrbuton ar k { ( k ) ( )} 2 k k ( x) E( x / x, x, x ) 2 ( x ) ( ) ( 2) k -- (4) V ( x / x, x, x ) E( x / x, x, x ) ( E( x / x, x, x )) 2 2 -- (5) whr k ( ) 2 { ( ) (2 x k k ) 2( )} 2 2 k E( x / x2, x3, x ) k ( x) ( ) ( 2)) k th Proof: Th n ordr momnt of th dstrbuton s 2 n n E( x / x, x, x ) x f ( x / x, x, x ) dx 0 k x {( ) ( )} n n k k ( ) 0 x 2 ( x ) E( x / x, x, x ) x dx ( ) ( 2) k ( nk ) ( x ) k { ( ){( n n ) ( )}} n k 2 k k ( x) E( x / x, x, x ) ( ) ( 2)) k 2 If n=, thn th Condtonal xctaton s k { ( k ) ( )} 2 k k ( x) E( x / x, x, x ) 2 ( x ) ( ) ( 2) k If n=2, thn th scond ordr momnt s k 2 { ( ) 2( k k ) 2( )} 2 2 k k ( x) E( x / x, x, x ) 2 ( x ) ( ) ( 2)) k Th condtonal varanc of th dstrbuton s obtand by Substtutng th frst and scond momnts n (5). Global Publshng Comany 4

G. S. DAVID SAM JAYAKUMAR, A. SOLAIRAJU, A. SULTHAN Thorm.5: If thr ar = (q + k) random varabls, such that q random varabls X, X, X, X X, X 2, X3, X q condtonally dnds on th k varabls,thn th dnsty functon of Sam-Sola s q q2 q3 qk multvarat addtv condtonal Gamma dstrbuton s ( x ) {( ) ( q k )} k q qk k q q q2 q3 qk k qk ( x ) f ( x, x, x, x / x, x, x, x ) ( ) ( k ) q k q x -- (6) whr 0, k 0 x Proof: Lt th multvarat condtonal law for q random varabls k varabls X, X, X, X q q2 q3 qk s gvn as X, X 2, X3, X q condtonally dndng on th f ( x, x, x, x / x, x, x, x ) q q q2 q3 qk f ( x, x, x, x, x, x, x, x ) q q q2 q3 qk f ( x, x, x, x ) q q2 q3 qk qk k x qk qk ( x ) {( ) ( q k )} k q q q2 q3 qk qk k x qk qk ( x ) q {( ) ( )} 0 0 0 k q k x q qk ( x ) {( ) ( q k )} k f ( x, x, x, x / x, x, x, x ) q q q2 q3 qk k qk ( x ) f ( x, x, x, x / x, x, x, x ) whr 0, k 0 x q k dx ( ) ( k ) q k Scton 2: Constants of Sam-Sola s multvarat addtv Gamma dstrbuton Thorm 2.: Th Margnal roduct momnts, Co-varanc and Poulaton Corrlaton Co-ffcnt btwn th Gamma random varabls X and X ar gvn as k E( x x ) k -- (7) ( k)( k2) COV ( x, x2) -- (8) ( k)( k2) ( x, x2) kk -- (9) whr ( x, x ) Global Publshng Comany 42

A NEW GENERALIZATION Proof: Assum that x and x 2 ar random varabls from Sam-Sola s multvarat addtv Gamma dstrbuton. Lt th roduct momnt of th dstrbuton s E( x x ) x x f ( x, x, x, x ) dx 0 0 0 Its Co-varanc s COV ( x, x ) E( x x ) E( x ) E( x ) -- (0) Thn k x ( x ) 0 0 0 k E( x x ) x x {( ) ( )} dx By valuaton, t follows that k E( x x ) k Th Margnal xctaton of Gamma varabls x and x 2 ar k / and k 2 / 2 rsctvly. Th Margnal Product momnt for E(x x 2 ) s obtand by substtutng th abov Margnal xctatons for x and x 2 n (0). Thus ( k)( k2) COV ( x, x2) -- () Corrlaton coffcnt of a dstrbuton s COV ( x, x2) ( x, x2) --(2a) It obsrvs that = k / and 2 = k 2 / 2 --(2b) From (), (2a) and (2b), t follows that ( k)( k2) ( x, x2) kk --(3) whr ( x, x ) Rmark 2.: Th Product momnts, Co-varanc and oulaton Corrlaton Co-ffcnt btwn th th and th j of Sam-Sola s multvarat addtv Gamma dstrbuton random varabl ar gvn as Global Publshng Comany 43

G. S. DAVID SAM JAYAKUMAR, A. SOLAIRAJU, A. SULTHAN k kj E( xx j) j ( k)( kj) COV ( x, x j) ( k)( kj) ( x, xj) kk j j --(4) --(5) --(6) whr j ; ( x, x ) Thorm 2.2: Th Momnt gnratng functon of Sam-Sola s Multvarat addtv Gamma dstrbuton s M t t t t k x, x2, x ( 3, x, 2, 3, ) ( ){( ( ) ) ( 2)} t t --(7) Proof: Lt th momnt gnratng functon of a Multvarat dstrbuton s gvn as tx M ( t, t, t t ) f ( x, x, x, x ) dx x, x2, x3, x, 0 0 0 tx k x ( ) x x, x2, x ( 3, x, 2, 3, ) {( ) ( )} 0 0 0 k M t t t t dx M t t t t k x, x2, x ( 3, x, 2, 3, ) ( ){( ( ) ) ( 2)} t t by ntgratng th abov quaton. Thorm 2.3: Th Cumulant of th Momnt gnratng functon of th Sam-Sola s Multvarat addtv Gamma dstrbuton s C t t t t k x, x2, x ( 3, x, 2, 3, ) log( ) log{( ( ) ) ( )} t t --(8) Proof: It s found from. C ( t, t, t t ) log( M ( t, t, t t )) x, x2, x3, x, x, x2, x3, x, Global Publshng Comany 44

A NEW GENERALIZATION Thorm 2.4: Th Charactrstc functon of th Sam-Sola s Multvarat addtv Gamma dstrbuton s j j k j ( t, t, t t ) ( ){( ( ) ) ( )} t t x, x2, x3, x, j j j j j j --(9) Proof: Lt th charactrstc functon of a multvarat dstrbuton s gvn as t x j j j x, x2, x ( 3, x t, t2, t3, t ) (, 2, 3, ) f x x x x dx j 0 0 0 j t x j j k j ( ) jxj j jx j j x, x2, x ( 3, x t, t2, t3, t ) {( ) ( )} j dx j 0 0 0 j j k j j j j k j ( t, t, t t ) ( ){( ( ) ) ( )} t t x, x2, x3, x, j j j j j j by ntgratng th abov quaton. Thorm 2.5: Th survval functon of th Sam-Sola s Multvarat addtv Gamma dstrbuton s (,, ) x x k S( x, x2, x3, x) ( )( ( ) ( )) x --(20) Proof: Lt th survval functon of a multvarat dstrbuton s gvn as S( x, x, x, x ) F( x, x, x, x ) x x2 x x 3 k ( ) u u 0 0 0 0 k S( x, x, x, x ) ( ( )) du du du du x k k u x u x k 0 S( x, x, x, x ) ( )( ( du ) ( )) (,, ) x x k S( x, x2, x3, x) ( )( ( ) ( )) x Whr varabl. x k k u u ( x,, k ) du k 0 s th lowr ncomlt Gamma ntgral of th random Global Publshng Comany 45

G. S. DAVID SAM JAYAKUMAR, A. SOLAIRAJU, A. SULTHAN Thorm 2.6: Th hazard functon of th Sam-Sola s Multvarat addtv Gamma dstrbuton s h( x, x, x, x ) k x x {( ) ( )} k (,, ) x x k x ( )( ( ) ( )) --(2) Proof: It s obtand from h( x, x, x, x ) f ( x, x, x, x ) S( x, x, x, x ) and S( x, x, x, x ) F( x, x, x, x ) Thorm 2.7: Th Cumulatv hazard functon of th Sam-Sola s Multvarat addtv Gamma dstrbuton s (,, ) x x k H( x, x2, x3, x) log( ( )( ( ) ( ))) x --(22) Proof: Lt th Cumulatv hazard functon of a multvarat dstrbuton s gvn as H( x, x, x, x ) log( F( x, x, x, x ) H( x, x, x, x ) log( S( x, x, x, x )) (,, ) x x k H( x, x2, x3, x) log( ( )( ( ) ( ))) x Scton 3: Som Scal Cass Rsult 3.: Th un-varat margnal of th Sam-Sola s multvarat addtv Gamma dstrbuton s th un-varat two aramtr Gamma dstrbutons. Rsult 3.2: If P = n (), th Sam-Sola s multvarat addtv Gamma dnsty s rducd to dnsty of un-varat two aramtr Gamma dstrbuton. Rsult 3.3: If P = 2 n (), thn th dnsty of Sam-Sola s Multvarat Gamma dstrbuton was rducd nto x x f ( x, x ) ( ) k k k2 k2 2 k k2 ( x x2 ) --(23) Global Publshng Comany 46

A NEW GENERALIZATION whr 0, x x2, k k2,,, 0 Ths s calld th dnsty of Sam-Sola s B-varat addtv Gamma dstrbuton. Rsult 3.4: Th tabls, tabl 2 and B-varat robablty surfac for (23) show th slctd smulatd standard Bvarat corrlatons btwn two Gamma random varabls whch ar boundd btwn - and + calculatd from 900 dffrnt combnatons of sha aramtr ( k, k ). Tabl : Smulaton runs for slctd valus of sha aramtr wth oulaton corrlaton bounds Runs 8 86 802 873 28 435 268 497 238 442 K.3 3 3.9 2 0.5 2 0.5.3 0. 0.4 K 2 3 2.3.9 3 2 0.5 0.5 2 0..3 2.7 533 2.7 0.4 ( x, x ) -0.9898-0.7539-0.50 +0.5 +0.7488 +0.984 Tabl 2: Smulaton runs and combnaton of sha aramtrs of Bvarat Gamma dstrbuton whn ( x, x2) =0 Runs K K 2 Runs K K 2 Runs K K 2 8 0.3 309 0.7 562.2 3. 3 0.5 577 0.8 67.6 350.9 589.3 76 0.5 363 0. 60 3 05. 372.6 686. 07 0.4 375 0.9 687 0.7 0 2.2 386.8 733.5 2.3 46.4 764.5 30 0.8 425 0.2 784.4 42 2.7 426 2. 79 0. 54 0.3 429 2.9 800.5 62.8 433 0.4 807.2 88.2 460.8 83.7 92 2.8 484 835.3 226.9 500 3 844.7 270.7 52 2.5 856 0.2 273.9 52 0.9 884 2 278 2.6 53.6 888.4 280 2.4 537 0.6 895 0.6 290.3 547 - - - Global Publshng Comany 47

G. S. DAVID SAM JAYAKUMAR, A. SOLAIRAJU, A. SULTHAN Fgur 2 Fg. Fg.2 Fg.3 Fg. 4 Fg.5 Fg.6 Global Publshng Comany 48

A NEW GENERALIZATION Fg. 7 Fg.8 Fg.9 Fg.0 Fg. Fg.2 Global Publshng Comany 49

G. S. DAVID SAM JAYAKUMAR, A. SOLAIRAJU, A. SULTHAN Rsult 3.6: From () and f k =, thn th oulaton corrlaton btwn th th and th j random varabl s gvn as ( x, xj) =0 and th dnsty of Sam-Sola s Multvarat addtv Gamma dstrbuton s rducd as roduct of unvarat xonntal dstrbuton, and ts dnsty functon s f ( x, x2, x3, x) ( ) x x f ( x, x, x, x ) -- (24) whr 0 x 0 Rsult 3.7: From() and f =, th Sam-sola s Multvarat addtv Gamma dstrbuton s rducd nto Sam- sola s Multvarat on aramtr addtv Gamma dstrbuton wth aramtrs as k and ts dnsty functon s gvn k x f ( x, x2, x3, x) {( ) ( )} k x -- (25) k whr 0 k 0 k x Rsult 3.8: From () and f and, thn th Sam-sola s Multvarat addtv Gamma dstrbuton s rducd nto Sam-sola s Multvarat two aramtr addtv Gamma dstrbuton wth aramtrs, k and ts dnsty functon s gvn as f x x x x x k k (, 2, 3, ) {( ) ( )} k x -- (26) Rsult 3.9: From () and f n, k 2 2 whr 0, k 0 dstrbuton s modfd nto Sam-sola s Multvarat addtv ch-squar and ts dnsty functon s gvn as x, thn th Sam-sola s Multvarat two aramtr addtv Gamma 2 -dstrbuton wth n dgrs frdom Global Publshng Comany 50

A NEW GENERALIZATION n x 2 ( ) f ( x 2, x2, x3, x) ( ) {( ) ( )} 2 n 2 2 x --(27) whr 0 x n 0 k Rsult 3.0: From () and f, thn th Sam-sola s Multvarat two aramtr addtv Gamma dstrbuton s altrd nto Sam-sola s Multvarat addtv Erlang-k dstrbuton and ts dnsty functon s gvn as k ( kx) f ( x, x, x, x ) ( k ){( ) ( )} k kx -- (28) y x whr 0 x, k 0 Rsult 3.: From () and f, thn th Sam-sola s Multvarat two aramtr addtv Gamma dstrbuton s transformd nto Sam-sola s Multvarat addtv Invrs Gamma dstrbuton and ts dnsty functon s gvn as k ( ) y f ( y, y, y, y ) ( ){( ) ( )} k 2 y ( ) y -- (29) Rsult 3.2: From () and f y x whr 0, k 0 y, thn th Sam-sola s Multvarat two aramtr addtv Gamma dstrbuton s transformd nto Sam-sola s Multvarat addtv log-gamma dstrbuton and ts dnsty functon s gvn as k log y ( log y ) f ( y, y2, y3, y ) ( ){( ) ( )} y k -- (30) Rsult 3.3: From () and fk m, whr m and y, k 0 y x, thn th Sam-sola s Multvarat two aramtr addtv Gamma dstrbuton s transformd nto Sam-sola s Multvarat Nagakam-m dstrbuton and ts dnsty functon s gvn as Global Publshng Comany 5

G. S. DAVID SAM JAYAKUMAR, A. SOLAIRAJU, A. SULTHAN m m ( ) 2 m y m y m y m m f ( y, y, y, y ) 2 ( ( )) -- (3) whr 0 y, 0 m 0.5 Concluson: Th multvarat gnralzaton of two aramtr Gamma dstrbuton s an addtv form of Sam- Sola s gnralzaton havng som ntrstng faturs. At frst, th margnal un-varat dstrbutons of th Sam- Sola s Multvarat addtv Gamma dstrbuton ar un-varat and njoyd th symmtrc rorty. Scondly, th Poulaton Corrlaton co-ffcnt of th roosd dstrbuton s boundd btwn - and + for crtan valus of sha aramtr and th authors stablshd th smulatd standard bvarat corrlatons. Thrdly, th Condtonal varanc of Sam-Sola s Multvarat addtv condtonal Gamma dstrbuton s htroscdastc n natur and ths fatur s a unqu for th roosd dstrbuton. Fnally, th multvarat gnralzaton of two aramtr Gamma dstrbuton n an addtv form on th way for th sam addtv form of th gnralzaton of th xonntal famly of Sam-Sola s Multvarat Ch-squar dstrbuton, Multvarat Erlang dstrbuton, Multvarat Invrs Gamma dstrbuton, Multvarat log Gamma dstrbuton and Multvarat Nagakam-m dstrbuton. REFERENCES: []Chryan, K. C. A bvarat corrlatd gamma-ty dstrbuton functon, Journal of th Indan Mathmatcal Socty, 5, (94), 33-44. [2]Krshnamoorthy, A. S., and Parthasarathy, M. A multvarat gamma- ty dstrbuton, Annals of Mathmatcal Statstcs, 22, 549-557. (95), 3, 229 [3]Ramabhadran,V. R.A multvarat gamma-ty dstrbuton, Sankhya,, (95), 45-46 [4]Krshnaah, P. R., and Rao, M. M. Rmarks on a multvarat gamma dstrbuton, Amrcan Mathmatcal Monthly, 68, (96), 342-346. [5]Sarmanov, I. O.A gnralzd symmtrc gamma corrlaton, Doklady Akadm Nauk SSSR, 79, 279-285; Sovt Mathmatcs Doklady, 9, (968), 547-550. [6]Gavr, D. P. Multvarat gamma dstrbutons gnratd by mxtur, Sankhya, Srs A, 32, (970), 23-26. [7]Johnson, N. L., and Kotz, S. Contnuous Multvarat Dstrbutons, frst dton, Nw York: John Wly &: Sons, 972. [8]Dussauchoy, A., and Brland, R. A multvarat gamma dstrbuton whos margnal laws ar gamma, In Statstcal Dstrbutons n Scntfc Work, Vol. (G. P. Patl, S. Kotz, and J. K. Ord, ds.), (975),- 39-328 [9]Bckr, P. J., and Roux, J. J. J. A bvarat xtnson of th gamma dstrbuton, South Afrcan Statstcal Journal, 5, (98), -2. [0]D'Est, G. M. A Morgnstrn-ty bvarat gamma dstrbuton, Bomtrka, 68, (98). 339-340. []Kowalczyk, T., and Tyrcha, J.Multvarat gamma dstrbutons, rorts and sha stmaton, Statstcs, J. Math. Bol. 26, 465-475, 20, (989).465-474. [2]Matha, A. M., and Moschooulos, P. G. On a multvarat gamma, Journal of Multvarat Analyss, 39, (99).35-53. [3]Matha, A. M., and Moschooulos, P. G. A form of multvarat gamma dstrbuton, Annals of th Insttut of Statstcal Mathmatcs, 44, (992), 97-06. Global Publshng Comany 52