1999 by CRC Press LLC

Σχετικά έγγραφα
p n r

Probability theory STATISTICAL MODELING OF MULTIVARIATE EXTREMES, FMSN15/MASM23 TABLE OF FORMULÆ. Basic probability theory

Homework for 1/27 Due 2/5


INTEGRATION OF THE NORMAL DISTRIBUTION CURVE

Solve the difference equation

1. For each of the following power series, find the interval of convergence and the radius of convergence:

n r f ( n-r ) () x g () r () x (1.1) = Σ g() x = Σ n f < -n+ r> g () r -n + r dx r dx n + ( -n,m) dx -n n+1 1 -n -1 + ( -n,n+1)

Solutions: Homework 3

Sixth Term Examination Papers MATHEMATICS LIST OF FORMULAE AND STATISTICAL TABLES

Introduction of Numerical Analysis #03 TAGAMI, Daisuke (IMI, Kyushu University)

DIPLOMA PROGRAMME MATHEMATICS SL INFORMATION BOOKLET

The Heisenberg Uncertainty Principle

Parameter Estimation Fitting Probability Distributions Bayesian Approach

LIST OF FORMULAE STATISTICAL TABLES MATHEMATICS. (List MF1) AND

Biorthogonal Wavelets and Filter Banks via PFFS. Multiresolution Analysis (MRA) subspaces V j, and wavelet subspaces W j. f X n f, τ n φ τ n φ.

Last Lecture. Biostatistics Statistical Inference Lecture 19 Likelihood Ratio Test. Example of Hypothesis Testing.

FREE VIBRATION OF A SINGLE-DEGREE-OF-FREEDOM SYSTEM Revision B

Presentation of complex number in Cartesian and polar coordinate system

Solution Series 9. i=1 x i and i=1 x i.

Lecture 17: Minimum Variance Unbiased (MVUB) Estimators

Ψηφιακή Επεξεργασία Εικόνας

IIT JEE (2013) (Trigonomtery 1) Solutions

Outline. M/M/1 Queue (infinite buffer) M/M/1/N (finite buffer) Networks of M/M/1 Queues M/G/1 Priority Queue

1. Matrix Algebra and Linear Economic Models

An Inventory of Continuous Distributions

Solutions to Exercise Sheet 5

Probability and Random Processes (Part II)

SUPERPOSITION, MEASUREMENT, NORMALIZATION, EXPECTATION VALUES. Reading: QM course packet Ch 5 up to 5.6

The Equivalence Theorem in Optimal Design

Matrices and Determinants

CHAPTER 103 EVEN AND ODD FUNCTIONS AND HALF-RANGE FOURIER SERIES

DERIVATION OF MILES EQUATION Revision D

HOMEWORK#1. t E(x) = 1 λ = (b) Find the median lifetime of a randomly selected light bulb. Answer:

MEI EXAMINATION FORMULAE AND TABLES (MF2)

α β

Diane Hu LDA for Audio Music April 12, 2010

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

LAD Estimation for Time Series Models With Finite and Infinite Variance

2. THEORY OF EQUATIONS. PREVIOUS EAMCET Bits.

Statistics 104: Quantitative Methods for Economics Formula and Theorem Review

Errata (Includes critical corrections only for the 1 st & 2 nd reprint)

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Μια εισαγωγή στα Μαθηματικά για Οικονομολόγους

2 Composition. Invertible Mappings

ΣΤΟΧΑΣΤΙΚΑ ΣΥΣΤΗΜΑΤΑ & ΕΠΙΚΟΙΝΩΝΙΕΣ 1o Τμήμα (Α - Κ): Αμφιθέατρο 4, Νέα Κτίρια ΣΗΜΜΥ Θεωρία Πιθανοτήτων & Στοχαστικές Ανελίξεις - 5

Outline. Detection Theory. Background. Background (Cont.)

k A = [k, k]( )[a 1, a 2 ] = [ka 1,ka 2 ] 4For the division of two intervals of confidence in R +

ANSWERSHEET (TOPIC = DIFFERENTIAL CALCULUS) COLLECTION #2. h 0 h h 0 h h 0 ( ) g k = g 0 + g 1 + g g 2009 =?

DERIVATION OF MILES EQUATION FOR AN APPLIED FORCE Revision C

On Generating Relations of Some Triple. Hypergeometric Functions

ST5224: Advanced Statistical Theory II

5.4 The Poisson Distribution.

3.4 SUM AND DIFFERENCE FORMULAS. NOTE: cos(α+β) cos α + cos β cos(α-β) cos α -cos β

Homework 3 Solutions

Appendix B: Mathematical Formulae and Statistical Tables

Section 8.3 Trigonometric Equations

On Inclusion Relation of Absolute Summability

CHAPTER 25 SOLVING EQUATIONS BY ITERATIVE METHODS

MATH 38061/MATH48061/MATH68061: MULTIVARIATE STATISTICS Solutions to Problems on Matrix Algebra

Vidyalankar. Vidyalankar S.E. Sem. III [BIOM] Applied Mathematics - III Prelim Question Paper Solution. 1 e = 1 1. f(t) =

Second Order Partial Differential Equations

Example Sheet 3 Solutions

Tired Waiting in Queues? Then get in line now to learn more about Queuing!

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

Homework 8 Model Solution Section

Areas and Lengths in Polar Coordinates

Chapter 5, 6 Multiple Random Variables ENCS Probability and Stochastic Processes

EE512: Error Control Coding

Quadratic Expressions

Factorial. Notations. Specific values. Traditional name. Traditional notation. Mathematica StandardForm notation. Specialized values

Degenerate Perturbation Theory

CS 1675 Introduction to Machine Learning Lecture 7. Density estimation. Milos Hauskrecht 5329 Sennott Square

Ordinal Arithmetic: Addition, Multiplication, Exponentiation and Limit

Bessel function for complex variable

APPENDICES APPENDIX A. STATISTICAL TABLES AND CHARTS 651 APPENDIX B. BIBLIOGRAPHY 677 APPENDIX C. ANSWERS TO SELECTED EXERCISES 679

Fibonacci. Notations. Primary definition. Specific values. Traditional name. Traditional notation. Mathematica StandardForm notation

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Inverse trigonometric functions & General Solution of Trigonometric Equations

SCHOOL OF MATHEMATICAL SCIENCES G11LMA Linear Mathematics Examination Solutions

Areas and Lengths in Polar Coordinates

Aquinas College. Edexcel Mathematical formulae and statistics tables DO NOT WRITE ON THIS BOOKLET

Answers - Worksheet A ALGEBRA PMT. 1 a = 7 b = 11 c = 1 3. e = 0.1 f = 0.3 g = 2 h = 10 i = 3 j = d = k = 3 1. = 1 or 0.5 l =

true value θ. Fisher information is meaningful for families of distribution which are regular: W (x) f(x θ)dx

Second Order RLC Filters

Lecture 3: Asymptotic Normality of M-estimators

Chapter 3: Ordinal Numbers

DESIGN OF MACHINERY SOLUTION MANUAL h in h 4 0.

Tridiagonal matrices. Gérard MEURANT. October, 2008

Math 6 SL Probability Distributions Practice Test Mark Scheme

( y) Partial Differential Equations

J. of Math. (PRC) Shannon-McMillan, , McMillan [2] Breiman [3] , Algoet Cover [10] AEP. P (X n m = x n m) = p m,n (x n m) > 0, x i X, 0 m i n. (1.

Srednicki Chapter 55

Other Test Constructions: Likelihood Ratio & Bayes Tests

Gaussian related distributions

Trigonometric Formula Sheet

STAT 330(Winter ) Mathematical Statistics

C.S. 430 Assignment 6, Sample Solutions

Bayesian statistics. DS GA 1002 Probability and Statistics for Data Science.

Fourier Series. constant. The ;east value of T>0 is called the period of f(x). f(x) is well defined and single valued periodic function

Transcript:

Poularias A. D. Probability ad Stochastic Processes The Hadboo of Formulas ad Tables for Sigal Processig. Ed. Aleader D. Poularias Boca Rato: CRC Press LLC,999 999 by CRC Press LLC

34 Probability ad Stochastic Processes 34. Aioms of Probability 34.. Aioms of Probability 34. Aioms of Probability 34. Coditioal Probabilities Idepedet Evets 34.3 Compoud (Combied)Eperimets 34.4 Radom Variable 34.5 Fuctios of Oe Radom Variable (r.v.) 34.6 Two Radom Variables 34.7 Fuctios of Two Radom Variables 34.8 Two Fuctios of Two Radom Variables 34.9 Epected Value, Momets, ad Characteristic Fuctio of Two Radom Variables 34. Mea Square Estimatio of R.V.'s 34. Normal Radom Variables 34. Characteristic Fuctios of Two Normal Radom Variables 34.3 Price Theorem for Two R.V.'s 34.4 Sequeces of Radom Variables 34.5 Geeral Cocepts of Stochastic Processes 34.6 Statioary Processes 34.7 Stochastic Processes ad Liear Determiistic Systems 34.8 Correlatio ad Power Spectrum of Statioary Processes 34.9 Liear Mea-Square Estimatio 34. The Filterig Problem for Statioary Processes 34. Harmoic Aalysis 34. Maroff Sequeces ad Processes Refereces I. PA ( ),II. PS ( ), III. If AB the PA ( + B) PA ( ) + PB ( )[. S a set of elemets of outcomes {ζ ι } of a eperimet (certai evet), empty set (impossible evet). {ζ ι } elemetary evet if {ζ ι } cosists of a sigle elemet. A + B uio of evets, AB itersectio of evets, evet a subset of S, P(A) probability of evet A. 999 by CRC Press LLC

34.. Corollaries of Probability Eample P( ), P( A) P( A),( A complemet set of A) PA ( + B) PA ( ) + PB ( ), PA ( + B) PA ( ) + PB ( ) PAB ( ) PA ( ) + PB ( ) S {hh,ht,th,tt} (tossig a coi twice), A {heads at first tossig} {hh,ht}, B {oly oe head came up} {ht,th}, G {heads came up at least oce} {hh,ht,th}, D {tails at secod tossig} {ht, tt} 34. Coditioal Probabilities Idepedet Evets 34.. Coditioal Probabilities PAM ( ) probability of evet AM PAM ( ) coditioal probaqbility of Agive M. PM ( ) probabilty of evet M.. 3. 4. PAM ( ) if AM PA PAM ( ) ( ) PM P ( A ) if ( ) AM A ( A M ) PAM PM ( ) ( ) if M A PM ( ) P( A+ BM) P( AM) + P( BM) if AB Eample P( f ) 6 /, i, L 6. M { odd} { f, f, f }, A { f }, AM { f }, P( M) 36 /, P( AM) 6 /,the i P( f eve) P( AM)/ P( M) 3 / 34.. Total Probability PB ( ) PBA ( ) PA ( ) + L + PBA ( ) PA ( ) arbitrary evet, AA i j,, L, A + L+ A S certai evet. 34..3 Baye's Theorem 3 5 PBA ( i) PA ( i) PA ( i B) PBA ( ) PA ( ) + + PBA ( ) PA ( ) L i j AA i j, i j,, L, A + A + LA S certai evet, Barbitrary 34..4 Idepedet Evets PAB ( ) PAPB ( ) ( ) implies A ad B are idepedet evets. 34..5 Properties. PAB ( ) PA ( ) 999 by CRC Press LLC

. 3. PBA ( ) PB ( ) PAA ( LA) PA ( ) LPA ( ), Ai idepedet evets 4. PA ( + B) PA ( ) + PB ( ) PAPB ( ) ( ) 5. AB ( A+ B), P( A+ B) P( A+ B), P ( AB) P( A) P( B) If A ad B are idepedet. Overbar meas complemet set. 6. If Ai A A3 are idepedet ad A is idepedet of AA 3 the PAAA ( 3) PA ( ) PA ( ) PA ( 3) PA ( ) PAA ( 3). Also PA [ ( A + A3)] PAA ( ) + PAA ( 3) PAAA ( 3) PA ( ) [ PA ( ) + PA ( ) PA ( ) PA ( )] PA ( ) PA ( + A) 3 3 3 34..6 PA ( + B+ C) PA ( ) + PB ( ) + PC ( ) PAB ( ) PAC ( ) PBC ( ) + PABC ( ) 34.3 Compoud (Combied, Eperimets 34.3. SS S Cartesia product Eample S {,,3}, S {heads, tails}, S S S {( heads),( tails),( heads),( tails),(3 heads),(3 tails)} 34.3. If A S, A S the A A ( A S )( A S ) (see Figure 34.) S A S A A A A S A S FIGURE 34. 34.3.3 Probability i Compoud Eperimets PA ( ) PA ( S) where ζ A ad ζ A 34.3.4 Idepedet Compoud Eperimets PA ( A) P( A) P( A) Eample P(heads) p, P(tails) q, p+ q, E eperimet tossig the coi twice E E (E eperimet of first tossig), E eperimet of secod tossig), S {h,t} P {h}p P {t}q, E E eperimet of the secod tossig, S S S [ hh, ht, th, tt}, P{ hh} P{} h P{} h p assume idepedece, P{ ht} pq, P{ th} qp, P[ t, t} q. For heads at the first tossig, H { hh, ht} or PH ( ) Phh { } + P{ ht} p + pq p 999 by CRC Press LLC

34.3.5 Sum of more Spaces S S + S, S outcomes of eperimet E ad S outcomes of eperimet E. S space of the eperimet E E + E; A A + A where A ad A are evets of E ad E : A S, A S ; PA ( ) PA ( ) + PA ( ). 34.3.6 Beroulli Trials PA ( ) probability of evet A, E E E... E perform eperimet times combied eperimet. p probability that evets occurs times i ay order pq ( ) PA ( ) ppa, ( ) qp, + q Eample A fair die was rolled 5 times. 5! 5 ( ) ( 5 )!! 6 6 probability that "four" will come up twice. Eample Two fair dice are tossed times. What is the probability that the dice total seve poits eactly four times? Solutio Evet 4 6 5 probability of B occurig four times ad B si times is. 543. 4 6 6 34.3.7 5 p 5 B P B 5 {( 6, ),( 5, ),( 34, ),( 43, ),( 5, ),( 6, )}, ( ) 6 p, P( 8) p 6 6 6 P { } probability of success of A (evet) will lie betwee ad P { } p( ) pq. The. Approimate value: pq p pq e pq ( ) /, >> πpq 34.3.8 DeMoivre-Laplace Theorem p pq pq e ( p)/ pq pq ( ), >> π 34.3.9 Poisso Theorem! pq!( )! ( p) a a e e p p a!!,,, p 999 by CRC Press LLC

34.3. Radom Poits i Time t t T t a T ( ) λt ( λ a ) a P { i ta} e e, t t ta << T, radom poits i (,T), λ / T. If!!, T, / T λ the approimatio becomes equality. ta. P{oe i ta} e λ λta λt P. lim { oe i t } a λ ta t 3. P ( ( t, t)} e 34.4 Radom Variables 34.4. Radom Variable a i t λ() t dt t To every outcome ζ of ay eperimet we assig a umber X( ζ ). The fuctio X, whose domai i the space S of all outcomes ad its rage is a set of umbers, is called a radom variable (r.v.). 34.4. Distributio Fuctio F ( ) P { X } defied o ay umber < <. { X } is a evet for ay real umber. 34.4.3 Properties of Distributio Fuctio. F( ), F( + ). F ( ) F ( ) for < 3. F ( + ) F ( ) cotiuous from the right 34.4.4 Desity Fuctio (or Frequecy Fuctio) a t t λ() tdt!, p α() t dt, α() t λ() t df( ) P { X + } f( ) ; f ( ) lim ; PX { } for cotiuous distributio fuctio; d o f( ) p δ i ( i) desity of discrete type, p F F i ( i) ( i ). i Eample Poisso distributio: PX { } λ λ e. The.!,,, L, λ > f( ) λ λ e δ! ( ) Eample If X is ormally distributed f e m ( ) / σ ( ) with m ad σ5, the the probability σ π m that X is betwee 9 ad,5 is P { X } f9ydy ) ( ) fydy ( ) ( ) + erf σ m erf erf + erf. 89 where error fuctio of erf y dy σ ep( / ) π t t 999 by CRC Press LLC

34.4.5 Tables of Distributio Fuctios (see Table 34.) TABLE 34. Distributio ad Related Quatities Defiitios. Distributio fuctio (or cumulative distributio fuctio [c.d.f.]): F ( ) probability that the variate taes values less tha or equal to P{ X } f( u) du. Probability desity fuctio (p.d.f.): f( ); P{ < X } f( ) d; f( ) l 3. Probability fuctio (discrete variates) f( ) probability that the variate taes the value P{ X } 4. Probability geeratig fuctio (discrete variates): u u l df d ( ) 5. Momet geeratig fuctio (m.f.g): Pt () Pt () t f( ), f( ) ( /!),,,,, X L > t t t µ rt Mt () t f( d ). Mt () + µ t+ + + µ L + L,! r! th r Mt () µ r r momet about the origi f( ) d r t M () t M () t M () t X+ Y X Y r r t 6. Laplace trasform of p.d.f.: 7. Characteristic fuctio : L s f () s e f( ) d, X Φ() t e jt f ( ) d, Φ () t Φ () t Φ () t X+ Y X Y 8. Cumulat fuctio: Kt () log Φ(), t KX+ Y() t KX() t + KY() t 9. r th r cumulat: the coefficiet of ( jt) / r! i the epasio of K(t). r th momet about the origi: r r Mt () µ r f( ) d r t r r Φ() t ( j) r t t t. Mea: µ first momet about the origi f ( ) d µ. r th momet about the mea: 3. Variace: σ secod momet about the mea ( µ ) f( ) d µ r µ r ( µ ) f ( ) d 4. Stadard deviatio: σ σ 5. Mea derivatio: µ f( ) d 6. Mode: A fractile (value of r.v.) for which the p.d.f is a local maimum 7. Media: m the fractile which is eceeded with probability /. 999 by CRC Press LLC

8. Stadardized r th momet about the mea: µ µ r ηr f( ) d r σ σ 3 9. Coefficiet of sewess: η µ / σ 3 3 4. Coefficit of urtois: η 4 µ 4 / σ. Coefficiet of variatio: (stadard deviatio) / mea. Iformatio cotet: I f( )log ( f( )) d 3. r th factorial momet about the origi (discrete case): r σ / µ r () µ ( r) Pt f( ) ( ) L( r + ), X, µ ( r) r t 4. r th factorial momet momet about the mea (discrete case): t 5. Relatioships betwee momets: µ ( r) f( µ )( µ )( µ ) L( µ r+ ), X µ r µ µ µ r µ µ i r r i( ); r r i( ), µ µ, µ i i i i 6. log is the atural logarithm Distributios r r v w v w. Beta: p.d.f f( ) ( ) / Bvw (, ), Bvw (, ) beta fuctio u ( u) du ;r th momet about the r origi ( v+ i)( v+ w+ i) ; mea v/( v+ w) ; variace vw/( v+ w) ( v+ w+ ) ; mode ( v ) /( v+ w+ ), v >, i / / w>; coefficiet of sewess: [ ( w v)( v+ w+ ) ]/[( v+ w+ )( vw) ] ; coefficiet of urtois: ([ 3( v+ w)( v+ w+ ) ( v+ )( w v)]/{ vw( v+ w+ )( v+ w+ 3 )]) + [ v( v w)]/( v+ w) ; coefficiet of variatio: [ w/[ v( v+ w+)]] / ; p.d.f. v w f( ) [( v+ w )! ( ) ]/[( v )!( w )!], v ad w itegers; Bvw (, ) Γ() vγ( w)/ Γ( v+ w) Bwv (, ), Γ() c ( c ) Γ( c ) v w4 v4 w p.d.f. f() vw v w. Biomial:, p is the umber of successes i idepedet Bemoulli trials where the probability of success at each trial is p ad the probability of failure is q p, positive iteger < p <. c.d.f i i F ( ) pq, i i 999 by CRC Press LLC

iteger; p.d.f. iteger; momet geeratig fuctio: ; probability geeratig f pq ( ), [ pep( t) + q] fuctio: ( pt + q) ; characteristic fuctio : Φ( t) [ p ep( jt) + q]. momets about the origi: meap, secod p(p + q), third p[( )( ) p + 3p( ) + ] ; momet about the mea: variace pq, third pq(q - p), fourth pq[ + 3pq( )] stadard deviatio : ( pq) / ; mode: p ( + ) p ( + ) ; coefficiet of sewess: ( q p) /( pq) / ; coefficiet of urtois: 3-(6/)+(/pq); factorial momets about the mea: secod pq, third pq( + q) ; coefficiet of variatio ( q/ p) / 3. Cauchy: p.d.f f( ) /[ πb[( a) / b] + ]], α shift parameter, b,scale parameter, < < ; mode a media a ( v )/ v/ 4. Chi-Squared: p.d.f. f( ) [ ep( / )]/[ Γ( v/ )], v (shape parameter) degrees of freedom, < ; v / v / momet geeratig fuctio : ( t), t > / ; characteristic fuctio: Φ( t) ( jt) ; cumulat fuctio: ; r th r cumulat; ; r th r ( v/ )log( jt) v[( r )!] momet about the origi: [ i+ ( v/ )] ; mea v; variace: v; stadard deviatio ( v ) / ; Laplace trasform of the p.d.f: ( + s) v / r i f(). F() v 4 5. Discrete uiform: a a+ b, iteger, a lower limit of the rage, b scale parameter; c.d.f F() ( a+)/ b ; p.d.f. f( ) / b ;probability geeratig fuctio: ( t a a b t )/( t) ;characteristic fuctio: ep[j(a-)t] sih( jtb / )sih( jt / ) / b ; mea: a+ ( b )/ ; variace: ( b )/ ; coefficiet of sewess ; iformatio cotet: log b. 6. Epoetial: <, b scale parameter mea, λ /b alterative parameter; c.d.f F( ) ep( / b) ; p.d.f f( ) ( / b)ep( / b) ; momet geeratig fuctio: /( bt), t > ( / b) ; Laplace trasform of the p.d.f: /( + bs) ; characteristic fuctio: /( jbt) ; cumulat fuctio: log( jbt ) ; r th cumulat: ( r )! b r ; r th momet about the origi: rb! r ; mea: b : variace: b ; stadard deviatio: b; mea deviatio: b/e (e base ad atural log); mode: ; media: b log ; coefficiet of sewess: ; coefficiet of urtosis 9; coefficiet of variatio: ; iformatio cotet: log ( eb ). 7. F-distributio: <, v ad w positive itegers degrees of freedom: p.d.f f( ) [ Γ[ ( v w)]( v/ w) v / + ; r th ( v )/ ( v+ w)/ ]/[ Γ( v) Γ ( w)( + v/ w) ] momet about the origi: [( w/ v) r Γ ( v r) Γ ( w r) /[ Γ ( v) Γ + ( w)], w > r ; mea: w/( w ), w > ; variace: [ w ( v+ w )]/[ v( w ) ( w 4)], w > 4 ; mode [ wv ( )]/[ vw ( + )], / / v > ;coefficiet of sewess: [( v+ w )[ 8( w 4)] ]/[( w 6)( v+ w ) ], w > 6;coefficiet of variatio: [[ ( v + / w ) /[ v9w 4)]], w > 4. 999 by CRC Press LLC

a8 f() f() v4 w3 4 4 8. Gamma: <, b scal e parameter > (or λ /b ), c> shaper parameter; p.d.f f( ) ( / b) c c c [ep( / b)]/[ bγ( c)], Γ( c) ep( u) u du ; momet geeratig fuctio: ( bt), t > / b ; Laplace frasform of the c c p.d.f.: ( + bs) ; characteristic fuctio: ( jbt) ; cumulat fuctio: c log 9 jbt) ; r th cumulat: ( r )! cb r ; r th r momet about the origi: b ( c+ i) ; mea: bc; variace: bc ; stadard deviatio: b c ; mode: b ( c ), c ; coefficiet of sewess: r i c / ; coefficiet of urtosis: 3 + 6/c: coefficiet of variatio: c / f() f() c / c / c c 4 4 9. Logoormal: <, m scale parameter media >, µ mea of log X >, m ep µ, µ log m, σ shape parameter stadard deviatio of log X, w ep( σ / ) ; p.d.f f( ) [ / σ( π) ]ep[ [log( / m)] / σ ] r th momet about the origi: m r ep( r σ / ) ; mea: m ep( σ / ) ; variace: mww ( ) : stadard deviatio: mw ( w) / ; mode m/w; media; m; coefficiet of sewess: ( w+ )( w ) / 4 3 ; coefficiet of urtosis: w + w + 3w 3 ; coefficiet of variatio: ( w ) /..8.4 f() m σ m σ.5 F() m σ.5 m σ 3 3. Negative bioomial: y umber of failures (iteger), umber of failures before success i a sequece of y + i i Beroulli trials; p probability of success at each trial, q p, y <, < p < ; c.d.f. Fy ( ) pq; i + y p.d.f. f y pq y ( ) ; momet geeratig fuctio: p ( qep t) ; probability geeratig fuctio: y p ( qt) ;; characteristic fuctio: p [ qep( jt)] ; cumulat fuctio: log( p) log( qep t); Cumulats: first 3 4 q / p, secod q / p, third q( + q)/ p, fourth q( 6q + p )/ p ;mea: q/p; Momets about the mea: variace 3 q / p, third q( + q)/ p, fourth ( q / p 4 )( q q p / 3 + 6 + ); stadard deviatio: ( q) / p; coefficiet of sewess: th i 999 by CRC Press LLC

/ 6 p ( + q)( q) ; coefficiet of urtosis: 3 + + factorial momet geeratig fuctio: r th factorial q ; ( t / ) q p momet about the origi: ( q/ p) r r / ( + r ) ; coefficiet of variato: ( q). f(y). p.5 f(y). 5 p.5 4 6 y 4 6 8 y. Normal: 8 < <, µ mea locatio parameter, σ stadard deviatio scale parameter, σ > ; p.d.f. / f( ) [ / σ( π ) ]ep[ ( µ ) / σ ]; momet geeratig fuctio: ep( µ t σ t ) ; characteristic fuctio: ep( jµ t σ t ); cumulat fuctio: jµ t σ t r th ; cumulat: K σ, K, r > ; mea: µ r th momet about r r/ the mea: µ r for r odd, µ ( σ r!) /[ [( r/ )!]] for r eve; variace: σ ; stadard deviatio: σ ; mea deviatio: σ( / π) / r / ; mode: µ; media: µ; coefficiet of sewess: ; coefficiet of urtosis: 3; iformatio cotet: log [ σ( πe) ] r + f() µ 3.5 F() 3 3. Pareto: <, c shape parameter; c.d.f. F ( ) c ; p.d.f. f c c ( ) ; r th momet about the origi: c/( c r), c > r; mea : c/( c ), c > ; variace: [ c/( c )] [ c/( c )], c > ; coefficiet of variatio: ( c )/[ c( c )] /, c >. f() c F() c 4 3. Pascal: umber of teals,, the Beroulli success parameter the umber of trials up to ad icludig the th success, p probability of success at each trial, < p <, q p; p.d.f. f momet geeratig pq ( ) ; fuctio: p ep( t) /( qep t) probability geeratig fuctio: ( pt) /( qt) ; characteristic fuctio: p ep( jt) /( q jt / ep( ) ; mea: /p; variace: q / p ; stadard deviatio: ( q) / p; coefficiet of variatio: ( q/ ) /. i 4. Poisso: <, λ mea (a parameter); c.d.f. F ( ) λ ep( λ ) / i!; p.d.f. momet geeratig fuctio: ep[ λ[ep( t) ]]; probability geeratig fuctio: ep[ λ( t)]; characteristic fuctio: i f( ) λ ep( λ) /!; 999 by CRC Press LLC

ep[ λ[ep( jt) ]] ;cumulat fuctio: λ[ep( t) ] t i / i!; r th cumulat: λ;momet about the origi: meaλ,secod i r λ + λ ; third λ[( λ + ) 3 r + λ],fourth λλ ( + 6λ + 7λ+ ); r th momet about the mea, µ i : λ µ i, r >, µ. i Momets about the mea: variace λ, thirdλ, fourth λ( + 3λ), fifth λ( + λ), sith λ( + 5λ + 5λ ); stadard deviatio λ / ;coefficiet of sewess: λ / ;coefficiet of urtosis: 3+ / λ ;factorial momets about the mea: secod λ, third -λ, fourth 3λ(λ+); coefficiet of variatio: λ /. i f().6 λ / f().3 λ f(). λ 6 4 4 6 4 6 8 5. Rectagular: a a+ b, rage, a lower limit,bscale parameter; c.d.f F ( ) ( a)/ b; p.d.f. f( ) / b; momet geeratig fuctio: ep( at)[ep( bt) ]/ bt; Laplace trasform of the p.d.f: ep( as)[ ep( bs)]/ bs; characteristic fuctio: ep( jat)[ep( jbt) ]/ jbt; mea: a+ b/ ; r th momet about the mea: µ r for r odd, µ r r ( b/ ) /( r+ ) for r eve;variace: b / ;stadard deviatio: b / ;mea deviatio b /4;media a+ b/ ;stadardized r th r / momet about the mea: µ r for r odd, µ r 3 /( r + ) for r eve; coefficiet of sewess: ; coefficiet of / urtosis: 95; coefficiet of variatio: b/[ 3 ( a+ b)] ; iformatio cotet: log b. f() F() b a a+b a a+b 6. Studet s: < <, v shape parameter (degrees of freedom), v positive iteger; p.d.f. f( ) [ Γ[( v+ )/ ] ( v+ )/ / [ + ( / v)] ]/[( πv) Γ ( v/ )] ; mea: ; r th momet about the mea: µ r for r odd, µ r / r [ 35 L( r ) v ]/ [( v )( v 4) L( v r)] for r eve, r<v: variace: v/( v ), v > ; mea deviatio: v / Γ( ( v )/ π / Γ( v); mode: ; coefficiet of sewess ad urtosis: f().5 v F().5 4 4 7. Weibull: <, b > scale parameter, c shape aprameter c>; c.d.f. F ( ) ep[ ( / b); p.d.f. f( ) ( c c / b c )ep[ ( / b) c ]; r th momet about the origi: b r Γ[( c+ r) / c]; mea: bγ[( c+ ) / c]. 999 by CRC Press LLC

f() c 3 b F() c 3 c c 3 3 TABLE 34. Normal Distributio Tables. / γ / f( ) distributio desity ( / π ) e, F( ) cumulative distributio fuctio ( / π ) τ, e d f ( ) f( ), f ( ) ( ) f( ), F( ) F ( ), P{ < X< } F ( ) F( ) f( ) f ( ) f ( ) F( ) f( ) f ( ) f ( )..5.3989 -. -.3989.5.695.35 -.76 -.64..54.3989 -. -.3989.5.695.353 -.787 -.59..58.3989 -.8 -.3987.5.6985.3485 -.8 -.543.3.5.3988 -. -.3984.53.79.3467 -.837 -.493.4.56.3986 -.59 -.398.54.754.3448 -.86 -.443.5.599.3984 -.99 -.3975.55.788.349 -.886 -.39.6.539.398 -.39 -.3968.56.73.34 -.9 -.34.7.579.398 -.79 -.396.57.757.339 -.933 -.89.8.539.3977 -.38 -.395.58.79.337 -.956 -.38.9.5359.3973 -.358 -.394.59.74.335 -.978 -.85..5398.397 -.99 -.3975.6.757.333 -.999 -.33..5438.3965 -.39 -.3968.6.79.33 -. -.8..5478.396 -.79 -.396.6.734.39 -.4 -.7.3.557.3956 -.38 -.395.63.7357.37 -.6 -.973.4.5557.395 -.358 -.394.64.7389.35 -.8 -.99.5.5596.3945 -.59 -.3856.65.74.33 -.99 -.865.6.5636.3939 -.63 -.3838.66.7454.39 -.8 -.8.7.5675.393 -.668 -.389.67.7486.387 -.36 -.757.8.574.395 -.77 -.3798.68.757.366 -.53 -.7.9.5753.398 -.744 -.3777.69.7549.344 -.7 -.647..5793.39 -.78 -.3754.7.758.33 -.86 -.593..583.39 -.8 -.373.7.76.3 -. -.538..587.3894 -.857 -.376.7.764.379 -.7 -.483.3.59.3885 -.894 -.368.73.7673.356 -.3 -.48.4.5948.3876 -.93 -.3653.74.774.334 -.45 -.373.5.5987.3867 -.967 -.365.75.7734.3 -.59 -.38.6.66.3857 -.3 -.3596.76.7764.989 -.7 -.6.7.664.3847 -.39 -.3566.77.7794.966 -.84 -.7.8.63.3836 -.74 -.3535.78.783.943 -.96 -.53.9.64.385 -.9 -.354.79.785.9 -.37 -.98.3.679.384 -.44 -.347.8.788.897 -.38 -.43.3.67.38 -.79 -.3437.8.79.874 -.38 -.988.3.655.379 -.3 -.34.8.7939.85 -.337 -.934.33.693.3778 -.47.3367.83.7967.87 -.346 -.88.34.633.3765.-.8 -.333.84.7995.83 -.355 -.85 999 by CRC Press LLC

TABLE 34. Normal Distributio Tables. (cotiued) / γ / f( ) distributio desity ( / π ) e, F( ) cumulative distributio fuctio ( / π ) e dτ, f ( ) f( ), f ( ) ( ) f( ), F( ) F ( ), P{ < X< } F ( ) F( ) f( ) f ( ) f ( ) F( ) f( ) f ( ) f ( ).35.6368.375 -.33 -.393.85.83.78 -.363 -.77.36.646.3739 -.346 -.355.86.85.756 -.37 -.78.37.6443.375 -.378 -.36.87.878.73 -.377 -.664.38.648.37 -.4 -.376.88.86.79 -.384 -.6.39.657.3697 -.44 -.335.89.833.685 -.389 -.558.4.6554.3683 -.473 -.394.9.859.66 -.395 -.56.4.659.3668 -.54 -.35.9.886.637 -.4 -.453.4.668.3653 -.534 -.38.9.8.63 -.44 -.4.43.6664.3637 -.564 -.965.93.838.589 -.48 -.35.44.67.36 -.593 -.9.94.864.565 -.4 -.99.45.6736.365 -.6 -.875.95.889 3.54 -.44 -.48.46.677.3589 -.65 -.83.96.835.56 -.46 -.97.47.688.357 -.679 -.783.97.834.49 -.47 -.47.48.6844.3555 -.77 -.736.98.8365.468 -.49 -.98.49.6879.3538 -.734 -.689.99.8389.444 -.4 -.49.5.695.35 -.76 -.64..843.4 -.4 -...843.4 -.4..5.933.95 -.943.69..8438.396 -.4.48.5.9345.76 -.97.633..846.37 -.49.96.5.9357.57 -.9.647.3.8485.37 -.48.43.53.937.38 -.894.66.4.858.33 -.46.9.54.938.9 -.877.67.5.853.99 -.44.36.55.9394. -.86.683.6.8554.75 -.4.8.56.946.8 -.843.694.7.8577.5 -.48.36.57.948.63 -.86.74.8.8599.7 -.45.37.58.949.45 -.89.74.9.86.3 -.4.44.59.944.7 -.79.7..8643.76 -.396.458.6.945.9 -.775.73..8665.55 -.39.5.6.9463.9 -.757.738..8686.3 -.386.54.6.9474.74 -.74.745.3.878.7 -.38.583.63.9484.57 -.73.75.4.879.83 -.375.64.64.9495.4 -.75.757.5.8749.59 -.368.664.65.955.3 -.687.76.6.877.36 -.36.74.66.955.6 -.67.766.7.879. -.354.74.67.955.989 -.65.77.8.88.989 -.347.78.68.9535.973 -.634.773.9.883.965 -.339.88.69.9545.957 -.67.776..8849.94 -.33.854.7.9554.94 -.599.778..8869.99 -.3.89.7.9564.95 -.58.779..8888.895 -.3.96.7.9573.99 -.563.78.3.897.87 -.33.96.73.958.893 -.546.78.4.895.849 -.93.994.74.959.878 -.58.78.5.8944.86 -.83.7.75.9599.863 -.5.78.6.896.84 -.73.6.76.968.848 -.49.778.7.898.78 -.6.9.77.966.833 -.474.777 999 by CRC Press LLC

TABLE 34. Normal Distributio Tables. (cotiued) / γ / f( ) distributio desity ( / π ) e, F( ) cumulative distributio fuctio ( / π ) e dτ, f ( ) f( ), f ( ) ( ) f( ), F( ) F ( ), P{ < X< } F ( ) F( ) f( ) f ( ) f ( ) F( ) f( ) f ( ) f ( ).8.8997.758 -.5.3.78.965.88 -.457.774.9.95.736 -.4.53.79.9633.84 -.439.77.3.93.74 -.8.8.8.964.79 -.4.769.3.949.69 -.4..8.9649.775 -.43.765.3.966.669 -.4.39.8.9556.76 -.386.76.33.98.647 -.9.67.83.9664.748 -.368.756.34.999.66 -.78.93.84.967.734 -.35.75.35.95.64 -.65.39.85.9678.7 -.333.746.36.93.58 -.5.344.86.9686.77 -.36.74.37.947.56 -.38.369.87.9693.694 -.98.734.38.96.539 -.5.39.88.9699.68 -.8.77.39.977.58 -..45.89.976.689 -.64.7.4.99.479 -.96.437.9.973.656 -.47.73.4.97.476 -.8.459.9.979.644 -.3.75.4.9.456 -.67.48.9.976.63 -.3.697.43.936.435 -.5.5.93.973.6 -.96.688.44.95.45 -.37.59.94.9738.68 -.79.679.45.965.394 -..537.95.9744.596 -.6.67.46.979.374 -.6.555.96.975.584 -.45.66.47.936.354 -.99.57.97.9756.573 -.9.65.48.936.334 -.975 -.588.98.976.56 -..64.49.939.35 -.959.64.99.9767.55 -.96.63.5.933.95 -.943.69..977.54 -.8.6..9773.54 -.8.6.5.9938.75 -.438.9..9778.59 -.64.69.5.994.7 -.49.96..9783.59 -.48.598.5.994.67 -.4.89.3.9788.58 -.3.586.53.9943.63 -.4.868.4.9793.498 -.6.575.54.9945.58 -.43.878.5.9798.488 -..563.55.9946.55 -.394.85.6.983.478 -.985.55.56.9948.5 -.386.836.7.989.468 -.969.538.57.9949.47 -.377.83.8.98.459 -.954.56.58.995.43 -.369.89.9.987.449 -.939.53.59.995.39 -.36.796..98.44 -.94.5.6.9953.36 -.353.78..986.43 -.99.487.6.9955.3.345.769..983.4 -.894.474.6.9956.9 -.338.756.3.9834.43 -.879.46.63.9957.6 -.33.743.4.9838.44 -.865.446.64.9959. -.33.73.5.984.396 -.85.433.65.996.9 -.36.77.6.9846.387 -.836.49.66.996.6 -.39.75.7.985.379 -.8.45.67.996.3 -.3.69.8.9854.37 -.88.39.68.9963. -.95.68.9.9857.363 -.794.377.69.9964.7 -.88.668 999 by CRC Press LLC

TABLE 34. Normal Distributio Tables. (cotiued) / γ / f( ) distributio desity ( / π ) e, F( ) cumulative distributio fuctio ( / π ) e dτ, f ( ) f( ), f ( ) ( ) f( ), F( ) F ( ), P{ < X< } F ( ) F( ) f( ) f ( ) f ( ) F( ) f( ) f ( ) f ( )..986.355 -.78.36.7.9965.4 -.8.656..9864.347 -.767.348.7.9966. -.75.644..9868.339 -.754.333.7.9967.99 -.69.63.3.987.33 -.74.39.73.9968.96 -.6.6.4.9875.35 -.77.34.74.9969.93 -.56.68.5.9868.37 -.74.89.75.997.9 -.5.597.6.988.3 -.7.75.76.997.88 -.44.585.7.9884.33 -.689.6.77.997.86 -.38.574.8.9887.97 -.676.45.78.9973.84 -.33.563.9.989.9 -.664.3.79.9974.8 -.7.56.3.9893.83 -.65.5.8.9974.79 -..54.3.9896.77 -.639..8.9975.77 -.6.53.3.9898.7 -.68.85.8.9976.75 -..5.33.99.64 -.66.7.83.9977.73 -.6.5.34.994.58 -.64.55.84.9977.7 -..5.35.996.5 -.593.4.85.9978.69 -.96.49.36.999.46 -.58.6.86.9979.67 -.9.48.37.99.4 -.57..87.9979.65 -.86.47.38.993.35 -.559.96.88.998.63 -.8.46.39.996.9 -.548.8.89.998.6 -.77.45.4.998.4 -.538.66.9.998.6 -.73.44.4.99.9 -.57.5.9.998.58 -.68.43.4.99.3 -.56.36.9.998.56 -.64.43.43.995.8 -.56..93.9983.55 -.6.44.44.997.3 -.496.7.94.9984.53 -.56.45.45.999.98 -.486.99.95.9984.5 -.5.396.46.993.94 -.476.978.96.9985.5 -.48.388.47.993.89 -.467.963.97.9985.48 -.44.379.48.9934.84 -.457.949.98.9986.47 -.4.37.49.9936.8 -.448.935.99.9986.46 -.37.363.5.9938.75 -.438.9 3..9987.44 -.33.355 3..9987.44 -.33.355 3.5.9998.9 -.3.98 3.5.9989.38 -.6.36 3.55.9998.7 -.6.85 3..999.33 -..8 3.6.9998.6 -..73 3.5.999.8 -.88.49 3.65.9999.5 -.9.63 3..9993.4 -.76. 3.7.9999.4 -.6.54 3.5.9994. -.66.94 3.75.9999.4 -.3.46 3.3.9995.7 -.57.7 3.8.9999.3 -..39 3.35.9996.5 -.49.49 3.85.9999. -.9.33 3.4.9997. -.4.3 3.9.. -.8.8 3.45.9997. -.36.3 3.95.. -.6.4 3.5.9998.9 -.3.98 4... -.5. 999 by CRC Press LLC

TABLE 34.3 Studet t-distributio Table f( ) + Γ y + / ) π Γ9 ( + )/ dy umber of degrees of freedom, umbers give of distributio, e.g., for 6 ad F.975,.447, F(-)-F() \F.6.75.9.95.975.99.995.9995.35. 3.78 6.34.76 3.8 63.657 636.69.89.86.886.9 4.33 6.965 9.95 3.598 3.77.765.638.353 3.8 4.54 5.84.94 4.7.74.533.3.776 3.747 4.64 8.6 5.67.77.476.5..57 3.365 4.3 6.869 6.65.78.44.943.447 3.43 3.77 5.959 7.63.7.45.895.365.998 3.499 5.48 8.6.76.397.86.36.896 3.555 5.4 9.6.73.383.833.6.8 3.5 4.78.6.7.37.8.8.764 3.69 4.587.6.697.363.796..78 3.6 4.437.59.695.356.78.79.68 3.55 4.38 3.59.694.35.77.6.65 3. 4. 4.58.69.345.76.45.64.977 4.4 5.58.69.34.753.3.6.947 4.73 6.58.69.337.746..583.9 4.5 7.57.689.333.74..567.898 3.965 8.57.688.33.734..55.878 3.9 9.57.688.38.79.93.539.86 3.883.57.687.35.75.86.58.845 3.85.57.686.33.7.8.58.83 3.89.56.686.3.77.74.58.89 3.79 3.56.685.39.74.69.5.87 3.767 4.56.685.38.7.64.49.797 3.745 5.56.684.36.78.6.485.787 3.75 6.56.684.35.76.56.479.779 3.77 7.56.684.34.73.5.473.77 3.69 8.56.683.33.7.48.467.763 3.674 9.56.683.3.699.45.46.756 3.659 3.56.683.3.697.4.457.75 3.646 4.55.68.33.684..43.74 3.55 6.54.679.96.67..39.66 3.46.54.677.89.658.98.358.67 3.373.53.674.8.645.96.36.576 3.9 34.4.6 Coditioal Distributio F ( M ) PX M P { X X M } {, }, PM { } { X, M} evet of all outcomes ζ such that X( ζ ) ad ζ M.. F( M), F( M) F M FM P X M P { < X ( ) ( ) { }, M } < PM { } 999 by CRC Press LLC

TABLE 34.4 The Chi-Squared Distributio y F ( ) ( )/ / y e dy / F ( / ) umber of degrees of freedom \F.5,.5.5..5.5.75.9.95.975.99.995.393.57.98.393..58..455.3.7 3.84 5. 6.63 7.88...56.3..575.39.77 4.6 6.5 7.38 9..6 3.77.5.6.35.584.584.37 4. 6.5 7.8 9.35.3.8 4.7.97.484.7.6.6.9 3.36 5.39 9.49. 3.3 4.9 5.4.554.83.5.6.67 4.35 6.63 9.4..8 5. 6.7 6.676.87.4.. 3.45 5.35 7.84.6.6 4.4 6.8 8.5 7.989.4.69.7.83 4.5 6.35 9.4. 4. 6. 8.5.3 8.34.65.8.73 3.49 5.7 7.34. 3.4 5.5 7.5.. 9.73.9.7 3.33 4.7 5.9 8.34.4 4.7 6.9 9..7 3.6.6.56 3.5 3.94 4.87 6.74 9.34.5 6. 8.3.5 3. 5..6 3.5 3.8 4.57 5.58 7.58.3 3.7 7.3 9.7.9 4.7 6.8 3.7 3.57 4.4 5.3 6.3 8.44.3 4,8 8.5. 3.3 6. 8.3 3 3.57 4. 5. 5.89 7.4 9.3.3 6. 9.8.4 4.7 7.7 9.8 4 4.7 4.66 5.63 6.57 7.79. 3.3 7.. 3.7 6. 9. 3.3 5 4.6 5.3 6.6 7.6 8.55. 4.3 8..3 5. 7.5 3.6 3.8 6 5.4 5.8 6.9 7.96 9.3.9 5.3 9.4 3.5 6.3 8.8 3. 34.3 7 5.7 6.4 7.56 8.67..8 6.3.5 4.8 7.6 3. 33.4 35.7 8 6.6 7. 8.3 9.39.9 3.7 7.3.6 6. 8.9 3.5 34.8 37. 9 6.84 7.63 8.9..7 4.6 8.3.7 7. 3. 3.9 36. 38.6 7.43 8.6 9.59.9.4 5.5 9.3 3.8 8.4 3.4 34. 37.6 4. 8.3 8.9.3.6 3. 6.3.3 4.9 9.6 3.7 35.5 38.9 4.4 8.64 9.94..3 4. 7..3 6. 3.8 33.9 36.8 4.3 4.8 3 9.6..7 3. 4.8 8..3 7. 3. 35. 38. 4.6 44. 4 9.89.9.4 3.8 5.7 9. 3.3 8. 33. 36.4 39.4 43. 45.6 5.5.5 3. 4.6 6.5 9.9 4.3 9.3 34.4 37.7 4.6 44.3 46.9 6.. 3.8 5.4 7.3.8 5.3 3.4 35.6 38.9 4.9 45.6 48.3 7.8.9 4.6 6. 8..7 6.3 3.5 36.7 4. 43. 47. 49.6 8.5 3.6 5.3 6.9 8.9.7 7.3 3.6 37.9 4.3 44.5 48.3 5. 9 3. 4.3 6. 7.7 9.8 3.6 8.3 33.7 39. 4.6 45.7 49.6 5.3 3 3.8 5. 6.8 8.5.6 4.5 9.3 34.8 4.3 43.8 47. 5.9 53.7 999 by CRC Press LLC

TABLE 34.5 The F-Distributio f r r Γ( r + r)/ ]( r / r) F( f) p{ F f} Γ( r / ) Γ( r / )[ + ( r / r )] PF { f} 95. / ( / ) ( r+ r)/ d r \r 3 4 5 6 7 8 9 5 4 3 4 6 6.4 99.5 5.7 4.6 3. 34. 36.8 38.9 4.5 4.9 43.9 45.9 48. 49. 5. 5. 6. 8.5 9. 9.6 9.5 9.3 9.33 9.35 9.37 9.38 9.4 9.4 9.43 9.45 9.45 9.46 9.47 9.48 3.3 9.55 9.8 9. 9. 8.94 8.89 8.85 8.8 8.79 8.74 8.7 8.66 8.64 8.6 8.59 8.57 4 7.7 6.94 6.59 6.39 6.6 6.6 6.9 6.4 6. 5.96 5.9 5.86 5.8 5.77 5.74 5.7 5.69 5 6.6 5.79 5.4 5.9 5.5 4.95 4.88 4.8 4.7 4.74 4.68 4.6 4.56 4.53 4.5 4.46 4.43 6 5.99 5.4 4.76 4.53 4.39 4.8 4. 4.5 4. 4.6 4. 3.94 3.87 3.84 3.8 3.77 3.74 7 5.59 4.74 4.35 4. 3.97 3.87 3.79 3.73 3.68 3.64 3.57 3.5 3.44 3.4 3.38 3.34 3.3 8 5.3 4.46 4.7 3.84 3.69 3.58 3.5 3.44 3.39 3.35 3.8 3. 3.5 3. 3.8 3.4 3. 9 5. 4.6 3.86 3.63 3.48 3.37 3.9 3.3 3.8 3.4 3.7 3..94.9.86.83.79 4.96 4. 3.7 3.48 3.33 3. 3.4 3.7 3..98.9.85.77.74.7.66.6 4.84 3.98 3.59 3.36 3. 3.9 3..95.9.85.79.7.65.6.57.53.49 4.75 3.89 3.49 3.6 3. 3..9.85.8.75.69.6.54.5.47.43.38 3 4.67 3.8 3.4 3.8 3.3.9.83.77.7.67.6.53.46,4.38.34.3 4 4.6 3.74 3.34 3..96.85.76.7.65.6.53.46.39.35.3.7. 5 4.54 3.68 3.9 3.6.9.79.7.64.59.54.48.4.33.9.5..6 6 4.49 3.63 3.4 3..85.74.66.59.37.49.4.35.8.4.9.5. 7 4.45 3.59 3..96.8.7.6.55.34.45.38.3.3.9.5..6 8 4.4 3.55 3.6.93.77.66.58.5.3.4.34.7.9.5..6. 9 4.38 3.5 3.3.9.74.63.54.48.3.38.3.3.6..7.3.98 4.35 3.49 3..87.7.6.5.45.39.35.8...8.4.99.95 4.3 3.47 3.7.84.68.57.49.4.37.3.5.8..5..96.9 4.3 3.44 3.5.8.66.55.46.4.34.3.3.5.7.3.98.94.89 3 4.8 3.4 3.3.8.64.53.44.37.3.7..3.5..96.9.86 4 4.6 3.4 3..78.6.5.4.36.3.5.8..3.98.94.89.84 999 by CRC Press LLC

5 4.4 3.39.99.76.6.49.4.34.8.4.6.9..96.9.87.8 6 4.3 3.37.98.74.59.47.39.3.7..5.7.99.95.9.85.8 7 4. 3.35.96.73.57.46.37.3.5..3.6.97.93.88.84.79 8. 3.34.95.7.56.45.36.9.4.9..4.96.9.87.8.77 9 4.8 3.33.93.7.55.43.35.8..8..3.94.9.85.8.75 3 4.7 3.3.9.69.53.4.33.7..6.9..93.89.84.79.74 4 4.8 3.3.84.6.45.34.5.8..8..9.84.79.74.69.64 6 4. 3.5.76.53.37.5.7..4.99.9.84.75.7.65.59.53 f r/ ( r/ ) Γ[( r + r) / ]( r / r) F( f) p{ F f} d F distributio ( r+ r)/ Γ( r / ) Γ( r / )[ + ( r / r )] PF { f}. 975 r \r 3 4 5 6 7 8 9 5 4 3 4 6 647.8 799.5 864. 899.6 9.8 937. 948. 956.7 963.6 968.6 976.7 984.9 993. 997. 6 38.5 39. 39.7 39.5 39.3 39.33 39.36 39.37 39.39 39.4 39.4 39.43 39.45 39.46 39.46 39.47 39.48 3 7.44 6.4 5.44 5. 4.88 4.73 4.6 4.54 4.47 4.4 4.34 4.5 4.7 4. 4.8 4.4 3.99 4..65 9.98 9.6 9.36 9. 9.7 8.98 8.9 8.84 8.85 8.66 8.56 8.5 8.46 8.4 8.36 5. 8.43 7.76 7.39 7.5 6.98 6.85 6.76 6.68 6.6 6.5 6.43 6.33 6.8 6.3 6.8 6. 6 8.8 7.6 6.6 6.3 5.99 5.8 5.7 5.6 5.5 5.46 5.37 5.7 5.7 5. 5.7 5. 4.96 7 87 6.54 5.89 5.5 5.9 5. 4.99 4.9 4.8 4.76 4.67 4.57 4.47 4.4 4.36 4.3 4.5 8 7.57 6.6 5.4 5.5 4.8 4.65 4.53 4.43 4.36 4.3 4. 4. 4. 3.95 3.89 3.84 3.78 9 5..7.8 4.7 4.48 4.3 4. 4. 4.3 3.96 3.87 3.77 3.67 3.6 3.56 3.5 3.45 6.94 5.46 4.83 4.47 4.4.7 3.95 3.85 3.78 3.7 3.6 3.5 3.4 3.37 3.3 3.6 3. 6.7 5.6 4.63 4.8 4.4 3.88 3.76 3.66 3.59 3.53 3.43 3.33 3.3 3.7 3. 3.6 3. 6.55 5. 4.47 4. 3.89 3.73 3.6 3.5 3.44 3.37 3.8 3.8 3.7 3..96.9.85 3 6.4 4.97 4.35 4. 3.77 3.6 3.48 3.39 3.3 3.5 3.5 3.5.95.89.84.78.7 4 6.3 4.86 4.4 3.89 3.66 3.5 3.38 3.9 3. 3.5 3.5.95.84.79.73.67.6 5 6. 4.77 4.5 3.8 3.58 3.4 3.9 3. 3. 3.6.96.86.76.7.64.59.5 6 6. 4.69 4.8 3.73 3.5 3.34 3. 3. 3.5.99.89.79.68.63.57.5.45 7 6.4 4.6 4. 3.66 3.44 3.8 3.6 3.6.98.9.8.7.6.56.5.44.38 8 5.98 4.56 3.95 3.6 3.38 3. 3. 3..93.87.77.67.56.5.44.38.3 9 5.9 4.5 3.9 3.56 3.33 3.7 3.5.96.88.8.7.6.5.45.39.33.7 999 by CRC Press LLC

TABLE 34.5 The F-Distributio f r r Γ( r + r)/ ]( r / r) F( f) p{ F f} Γ( r / ) Γ( r / )[ + ( r / r )] PF { f} 95. / ( / ) ( r+ r)/ d r \r 3 4 5 6 7 8 9 5 4 3 4 6 5.87 4.46 3.86 3.5 3.9 3.3 3..9.84.77.68.57.46.4.35.9. 5.83 4.4 3.8 3.48 3.5 3.9.97.87.8.73.64.53.4.37.3.5.8 5.79 4.38 3.78 3.44 3. 3.5.93.84.76.7.6.5.39.33.7..4 3 5.69 4.9 3.69 3.35 3.3.97.85.75.68.6.5.4.3.4.8..5 4 5.7 4.3 3.7 3.38 3.5.99.87.78.7.64.54.44.33.7..5.8 5 5.69 4.9 3.69 3.35 3.3.97.85.75.68.6.5.4.3.4.8..5 6 5.66 4.7 3.67 3.33 3..94.8.73.65.59.49.39.8..6.9.3 7 5.63 4.4 3.65 3.3 3.8.9.8.7.63.57.47.36.5.9.3.7. 8 5.6 4. 3.63 3.9 3.6.9.78.69.6.55.45.34.3.7..5.98 9 5.59 4. 3.6 3.7 3.4.88.76.67.59.53.43.3..5.9.3.96 3 5.57 4.8 3.59 3.5 3.3.87.75.65.57.5.4.3..4.7..94 4 5.4 4.5 3.46 3.3.9.74.6.53.45.39.9.8.7..94.88.8 6 5.9 3.93 3.34 3..79.63.5.4.33.7.7.6.94.88.8.74.67 F( f) p{ F f} PF { f} 99. f r / ( r / ) ( r+ r)/ Γ[( r + r) / ]( r / r) Γ( r / ) Γ( r / )[ + ( r / r )] d r \r 3 4 5 6 7 8 9 5 4 3 4 6 45 4999.5 543 565 5764 5859 598 598 6 656 66 657 69 635 66 687 633 98.5 99. 99.7 99.5 99.3 99.33 99.36 99.37 99.39 99.4 99.4 99.43 99.45 99.46 99.47 99.47 99.48 3 34. 3.8 9.46 8.7 8.4 7.9 7.67 7.49 7.35 7.3 7.5 6.87 6.69 6.6 6.5 6.4 6.3 4. 8. 6.69 5.98 5.5 5. 4.98 4.8 4.66 4.55 4.37 4. 4. 3.93 3.84 3.75 3.65 5 6.6 3.7.6.39.97.67.46.9.6.5 9.89 9.7 9.55 9.47 9.38 9.9 9. 6 3.75.9 9.78 9.5 8.75 8.47 8.6 8. 7.98 7.87 7.7 7.56 7.4 7.3 7.3 7.4 7.6 7.5 9.55 8.45 7.85 7.46 7.9 6.99 6.84 6.7 6.6 6.47 6.3 6.6 6.7 5.99 5.9 5.8 999 by CRC Press LLC

8.6 8.65 7.59 7. 6.63 6.37 6.8 6.3 5.9 5.8 5.67 5.5 5.36 5.8 5. 5. 5.3 9.56 8. 6.99 6.4 6.6 5.8 5.6 5.47 5.35 5.6 5. 4.96 4.8 4.73 4.65 4.57 4.48.4 7.56 6.55 5.99 5.64 5.39 5. 5.6 4.94 4.85 4.7 4.56 4.4 4.33 4.5 4.7 4.8 9.65 7. 6. 5.67 5.3 5.7 4.89 4.74 4.63 4.54 4.4 4.5 4. 4. 4. 3.94 3.86 9.33 6.93 5.95 5.4 5.6 4.8 4.64 4.5 4.39 4.3 4.6 4. 3.86 3.78 3.7 3.6 3.54 3 9.7 6.7 5.74 5. 4.86 4.6 4.44 4.3 4.9 4. 3.96 3.8 3.66 3.59 3.5 3.43 3.34 4 8.86 6.5 5.56 5.4 4.69 4.46 4.8 4.4 4.3 3.94 3.8 3.66 3.5 3.43 3.35 3.7 3.8 5 8.68 6.36 5.4 4.89 4.56 4.3 4.4 4. 3.89 3.8 3.67 3.5 3.37 3.9 3. 3.3 3.5 6 8.53 6.3 5.9 4.77 4.44 4. 4.3 3.89 3.78 3.69 3.55 3.4 3.6 3.8 3. 3..93 7 8.4 6. 5.8 4.67 4.34 4. 3.93 3.79 3.68 3.59 3.46 3.3 3.6 3.8 3..9.83 8 8.9 6. 5.9 4.58 4.5 4. 3.84 3.7 3.6 3.5 3.96 3.3 3.8 3..9.84.75 9 8.8 5.93 5. 4.5 4.7 3.94 3.77 3.63 3.3 3.43 3.8 3.5 3..9.84.76.67 8. 5.85 4.94 4.43 4. 3.87 3.7 3.56 3.46 3.37 3.3 3.9.94.86.78.69.6 8. 5.78 4.87 4.37 4.4 3.8 3.64 3.5 3.4 3.3 3.7 3.3.88.8.7.64.55 7.95 5.7 4.8 4.3 3.99 3.76.59 3.45 3.35 3.6..98.83.75.67.58.83 3 7.88 5.66 4.76 4.6 3.94 3.7 3.54 3.4 3.3 3. 3.7.93.78.7.6.54.75 4 7.8 5.6 4.7 4. 3.9 3.67 3.77 3.36 3.6 3.7 3.3.89.74.66.58.49.67 5 7.77 5.57 4.68 4.8 3.85 3.63 3.46 3.3 3. 3.3.99.85.7.6.54.45.36 6 7.7 5.53 4.64 4.4 3.8 3.59 3.4 3.9.8 3.9.96.8.66.58.5.4.33 7 7.68 5.49 4.6 4. 3.78 3.56 3.39 3.6 3.5 3.6.93.78.63.55.47.38.9 8 7.64 5.45 4.57 4.7 3.75 3.53 3.36 3.3 3. 3.3.9.75.6.5.44.35.6 9 7.6 5.4 4.54 4.4 3.73 3.5 3.33 3. 3.9 3..87.73.57.49.4.33.3 3 7.56 5.39 4.5 4. 3.7 3.47 3.3.3.7 3.7.98.84.7.55.47.39.3. 4 7.3 5.8 4.3 3.83 3.5 3.9 3..99.89.8.66.5.37.9... 6 7.8 4.98 4.3 3.65 3.34 3..95.8.7.63.5.35...3.94.84 f r/ ( r/ ) Γ[( r + r) / ]( r / r) F( f) p{ F f} d ( r+ r)/ Γ( r / ) Γ( r / )[ + ( r / r )] PF { f}. 995 r \r 3 4 5 6 7 8 9 5 4 3 4 6 6 65 5 356 3437 375 395 49 44 446 463 4836 49 544 548 553 999 by CRC Press LLC

TABLE 34.5 The F-Distributio (cotiued) f r r Γ( r + r)/ ]( r / r) F( f) p{ F f} Γ( r / ) Γ( r / )[ + ( r / r )] PF { f}. 995 / ( / ) ( r+ r)/ d r \r 3 4 5 6 7 8 9 5 4 3 4 6 98.5 99. 99. 99. 99.3 99.3 99.4 99.4 99.4 99.4 99.4 99.4 99.4 99.5 99.5 99.5 99.5 3 55.55 49.8 47.47 46.9 45.39 44.84 44.43 44.3 43.88 43.69 43.39 43.8 4.78 4.6 4.47 4.3 4.5 4 3.33 6.8 4.6 3.5.46.87.6.35.4.97.7.44.7.3 9.89 9.75 9.6 5.78 8.3 6.53 5.56 4.94 4.5 4. 3.96 3.77 3.6 3.38 3.5.9.78.66.53.4 6 8.63 4.54.9.3.46.7.79.57.39.5.3 9.8 9.59 9.47 9.36 9.4 9. 7 6.4.4.88.5 9.5 9.6 8.89 8.68 8.5 8.38 8.8 7.97 7.75 7.65 7.53 7.4 7.3 8 4.69.4 9.6 8.8 8.3 7.95 7.69 7.5 7.34 7. 7. 6.8 6.6 6.5 6.4 6.9 6.8 9 3.6. 8.7 7.96 7.47 7.3 6.88 6.69 6.54 6.4 6.3 6.3 5.83 5.73 5.6 5.5 5.4.83 9.43 8.8 7.34 6.87 6.54 6.3 6. 5.97 5.85 5.66 5.47 5.7 5.7 5.7 4.97 4.86.3 8.9 7.6 6.88 6.4 6. 5.86 5.68 5.54 5.4 5.4 5.5 4.86 4.76 4.65 4.55 4.44.75 8.5 7.3 6.5 6.7 5.76 5.5 5.35 5. 5.9 4.9 4.7 4.53 4.43 4.33 4.3 4. 3.37 8.98 6.93 6.3 5.79 5.48 5.5 5.8 4.94 4.8 4.64 4.46 4.7 4.7 4.7 3.97 3.87 4.6 7.9 6.68 6. 5.56 5.6 5.3 4.86 4.7 4.6 4.43 4.5 4.6 3.96 3.86 3.76 3.66 5.8 7.7 6.48 5.8 5.37 5.7 4.85 4.67 4.54 4.4 4.5 4.7 3.88 3.79 3.69 3.58 3.48 6.58 7.5 6.3 5.64 5. 4.9 4.69 4.5 4.38 4.7 4. 3.9 3.73 3.64 3.54 3.44 3.33 7.38 7.35 6.6 5.5 5.7 4.78 4.56 4.39 4.5 4.4 3.97 3.79 3,6 3.5 3.4 3.3 3. 8. 7. 6.3 5.37 4.96 4.66 4.44 4.8 4.4 4.3 3.86 3.68 3.5 3.4 3.3 3. 3. 9.7 7.9 5.9 5.7 4.85 4.56 4.34 4.8 4.4 3.93 3.76 3.59 3.4 3.3 3. 3. 3. 9.94 6.99 5.8 5.7 4.76 4.47 4.6 4.9 3.96 3.85 3.68 3.5 3.3 3. 3. 3..9 9.83 6.89 5.73 5.9 4.68 4.39 4.8 4. 3.88 3.77 3.6 3.43 3.4 3.5 3.5.95.84 9.73 6.8 5.65 5. 4.6 4.3 4. 3.94 3.8 3.7 3.54 3.36 3.8 3.8.98.88.77 3 9.63 6.73 5.58 4.95 4.54 4.6 4.5 3.88 3.75 3.64 3.47 3.3 3. 3..9.8.7 4 9.55 6.66 5.5 4.89 4.49 4. 3.99 3.83 3.69 3.59 3.4 3.5 3.6.97.87.77.66 999 by CRC Press LLC

5 9.48 6.6 5.46 4.84 4.43 4.5 3.94 3.78 3.64 3.54 3.37 3. 3..9.8.7.6 6 9.4 6.54 5.4 4.79 4.38 4. 3.89 3.73 3.6 3.49 3.33 3.5.97.87.77.67.84 7 9.34 6.49 5.36 4.74 4.34 4.6 3.85 3.69 3.56 3.45 3.8 3..93.83.73.63.77 8 9.8 6.44 5.3 4.7 4.3 4. 3.8 3.65 3.5 3.4 3.5 3.7.89.79.69.59.7 9 9.3 6.4 5.8 4.66 4.6 3.98 3.77 3.6 6.48 3.48 3. 3.4.86.76.66.56.66 3 9.8 6.35 5.4 4.6 4.3 3.95 3.74 3.58 3.45 3.34 3.8 3..8.73.63.5.4 4 8.83 6.7 4.98 4.37 3.99 3.7 3.5 3.35 3. 3..95.78.6.5.4.3.8 6 8.49 5.79 4.73 4.4 3.76 3.49 3.9 3.3 3..9.74.57.39.9.9.8.96 999 by CRC Press LLC

TABLE 34.6 The Poisso fuctio e f( )! λ λ 999 by CRC Press LLC

TABLE 34.7 The Poisso Distributio e F ( ) λ λ! 999 by CRC Press LLC

34.4.7 Coditioal Desity f M df ( M ) P X M ( ) lim { + } d f ( M ) d F ( M ) Eample X( fi ) i, i, L6 where f i face of a die. M { f, f4, f6 ] eve evet. For 6, { X, M} {,, } { f, f, f }, f( M) F f f f 4 6 4 6 ; for 4 <6, { X, M} { f, f4}, F( M) P{ f, f4}/ PM { } PM { } ( / 6)/( 3/ 6) / 3; for < 4,{ X, M} { f}, F( M) P{ f}/ P{ M} (/6)(3/6) /3; for <, { X, M} ad FM ( ). 34.4.8 Total Probability F ( ) FA ( ) PA ( ) + FA ( ) PA ( ) + + FA ( ) PA ( ) PA ( ), L their sum is equal to the certai evet S. A i s are mutually eclusive ad 34.5 Fuctio of Oe Radom Variable (r.v.) 34.5. Radom Variable (Defiitio) To every eperimetal outcome ζ we assig a umber rage is the set I X of the real umbers X( ζ ).. X( ζ ). The domai of X is the space S, ad its 34.5. Fuctio of r.v. Y g( X) g[ X( ζ )] 34.5.3 Distributio Fuctio of Y (see 34.5.) F ( y) P{ Y y} P{ g( X) y} P{ X I } y y Note: To fid F ( y ) for a give y we must fid that set I y y ad the probability that X is i I y.refer to Figure 34.:If y the g ( ) yfor ay. Hece { Y y} certai evet ad F ( y ) P { Y y } y.if y y, the g ( ) y for ad,hece, Fy( y) P{ Y y} P{ X } F( ) ( depeds o y ), If y y, the g ( ) y has three solutios,, : g ( ) g ( ) g ( ) y ad from Figure 34. g ( ) y if or ad hece, F ( y ) P { X } P { } Y + FX( ) + FX( ) FX( ). If y < l o value of produces g ( ) y ad the evet { Y y} has zero probability: Fy( y). Eample Y / X. If y >, there are two solutios: y, / y. g( ) y if or ad thus F ( y) P{ Y y} P{ X / y} + P{ X / y} F ( / y) + F ( / y). y if y <, o will produce g ( ) y ad, hece, F ( y y ). 999 by CRC Press LLC

FIGURE 34. 34.5.4 Desity Fuctio of Yg(X) i Terms of f X () of X ) Solve yg() for i terms of y. If,, L, are all its real roots, the y g ( g ) L ( ) L, fx ( ) fx( ) the fy ( y) + L+ + L, g ( ) dg( )/ d. If yg() has o real roots the fy y. g ( ) g ( ) ( ) Eample g( ) ax + b ad ( y b)/ a for every y. g ( ) a ad hece f ( y) Y a f y b X a Eample gx ( ) ax with the r.v. y a, a>. If y < roots are imagiary ad f ( y ) Y. If y > the y / a ad y/ a.sice g ( ) a ay ad g ( ) a ay,the f Y (y) fx ay Eample 3 y f a + X y uy uy a ( ), ( ) uit step fuctio. Y asi( X + θ), a>. If y < a the y asi( + θ) has ifiitely may solutios L,,,, L. dg( ) / d acos( + θ) a y ad from 34.5.4 f ( y) / a y f ( ), y < a. For y > there eist o solutios, ad f ( Y y ).; y si a θ, Eample 4 ax Y be u( X), a>, b>. If y < or y > b the the equatio y bep( a) u( ) has o solutio, ad hece f y If the ad Y y b, ( / a) l( y / b). g ( ) abe ay f Y y f ( ( / a) l( y/ b)) / ay, y b.. X Y ( ) X 999 by CRC Press LLC

34.5.5 Coditioal Desity of Yg() f Y ( ym) fx ( M) fx( M) + L+ + L g ( ) g ( ) Eample fx ( ) Y ax, a>, X, f ( X ) u ( ) (see 34.5.4 Eample ), ad hece fy yx F ( ) ( ) fx ( y/ a) [ /( ay)] u ( ). F ( ) X X f( X t) f( )/{ F( t)] f( ) d, t t 34.5.6 Epected Value E{ X} f ( ) d cotiuous r.v. EX { } PX { } p discrete r.v. 34.5.7 Epected Value of a Fuctio g(x) E{ Y g( X)} yf ( y) dy g( ) f ( ) d Y cotiuous r.v. EgX {( )} g ( ) PX { } discrete type of r.v. 34.5.8 Coditioal Epected Value EXM { } f( Md ) cotiuous r.v. EXM { } PX { M} discrete r.v. 34.5.9 Variace σ E{( X µ ) ( µ ) f( ) d cotiuous r.v. σ ( µ ) P{ X } discrete r.v. 999 by CRC Press LLC

σ EX { } E{ X} Eample PX { ) e Poisso distributio.!, L,, but or λ λ ad hece, EX { } λ.. 34.5. Momets About the Origi 34.5. Cetral Momets 34.5. Absolute Momets 34.5.3 Geeralized Momets λ λ λ λ λ λ EX { ) e e e.!!! d d e λ d λ λ e λ dλ! λ! λ λ λ e! r µ EX { } f( d ) µ µ r, µ µ EX { }, µ r r r r r r r µ EX µ f d E µ X µ µ r r r { } ( ) ( ) ( ) r ( ) 3 µ µ, µ µ µ, µ µ µµ + µ µ µ, µ µ 3µµ + 3µ µ µ µ 3 3µµ + µ M E{ X } f( ) d a E X a am E X a µ {( ) }, { } λ 3 3 r 3 999 by CRC Press LLC

Eample a a a EX { } d, E { } a σ a + 3 for X uiformly distributed i (-a,a). Eample b+ a EX { } Γ( b + ) b a e b+ a Γ( b+ + ) d b+ + a Γ( b+ ) for a gamma desity b b a f( ) { a + / Γ( b+ )] e u ( ), u ( ) uit step fuctio. 34.5.4 Tchebycheff Iequality σ P{ X µ σ }, µ EX { }. Regardless of the shape of f( ), P{ µ ε < X < µ + ε} ε Geeralizatios:. If f ( y y ) the PY EY {} { α }, α > α. E{ X α } P{ X α ε } ε 34.5.5 Characteristic Fuctio Φ( ), Φ( ω) Eample jω Φ( ω) Ee { } f( d ) for cotiuous r.v. jω Φ( ω) e P{ X } for discrete type r.v. jωy jω( ax+ b) jωb jωax Φ( ω) Ee { } Ee { } e Ee { }, Eample if Y ax + b PX { } λ λ e Poisso distributio!, L,, λ jωλ λ jω Φ( ω) e e e ( e )! 34.5.6 Secod Characteristic Fuctio Ψ( ω) l Φ( ω) 999 by CRC Press LLC

34.5.7 Iverse of the Characteristic Fuctio jω f( ) Φ( ω) e dω π 34.5.8 Momet Theorem ad Characteristic Fuctio d Φ( ) j µ E X, { } dω 34.5.9 Covolutio ad Characteristic Fuctio Φ( ω) Φ ( ω) Φ ( ω), where Φ ( ω) ad Φ ( ω) are the characteristic fuctios of the desity fuctios f ad. ω ( + ) ω Ee j X X } ad ( ) f f ( ) ( ) where idicates covolutio. 34.5. Characteristic Fuctio of Normal r.v. Φ( ω ) ep( jµω σ ω ) 34.6 Two Radom Variables 34.6. Joit Distributio Fuctio F ( y) P{ X, Y y}, F (, ) F ( ), F (, y) F ( y), y y y y F (, ), F (, y), F (, ) y y y 34.6. Joit Desity Fuctio Fy (, ) f(, y), f ( ) f (, y ) dy, fy ( y ) f (, y ) d y 34.6.3 Coditioal Distributio Fuctio PY { ym, } PX { Y, y} Fy(, y) Fy ( ym) P{ Y ym}, FyX { } PM { } PX { } F ( ) PX { ay, by, y} y b Fy ( yx a, Y b) PX { ay, b} Fy( a, y)/ Fy( a, b) y< b 999 by CRC Press LLC

34.6.4 Coditioal Desity Fuctio F y y f y d f y d y(, )/ y( ξ, ) ξ y(, ) fy( yx ), fy( y < X F ( ) f (, y) d dy ( ) F( ), ξ ξ Γ fy(, y) fy( yx ) f ( ) 34.6.5 Baye's Theorem y f ( yx ) y f ( Y y) f ( y) f ( ) y 34.6.6 Joit Coditioal Distributio F (, ya< X b) y Fy( b, y) Fy( a, y) F( b) F( a) PX { Y, ya, < Y b} Fy(, y) Fy( a, y) Pa { < X b} F( b) F( a) > b a< b a 34.6.7 Coditioal Epected Value gyf ( ) y( ydy, ) EgY {( ) X } gy () fy( yx dy ), EEYX { { }} EY { } f (, y) dy 34.6.8 Idepedet r.v. 34.6.9 Joitly Normal r.v. EX { } µ, EY { } µ, σ σ, σ y σ. If r, f(, y) f( ) fy( y) idepedet. r <, r correlatio coeffifiet. F (, y) F ( ) F ( y); f (, y) f( ) f( y); f ( y) f ( y); f ( y) f ( ) y y y y y ( µ ) r( )( y ) ( y ) µ µ µ f(, y) ep + πσ σ r ( r ) σ σσ σ y 999 by CRC Press LLC

Coditioal Desities rσ fy( yx ) ep y µ ( ) µ σ ( r ) ( r ) π σ σ rσ EYX { } µ + ( µ ), σ σ r y σ If r µ µ the EY { X } ( r σ σ ) + σ 34.7 Fuctios of Two Radom Variables 34.7. Defiitios Z g( X, Y) g[ X( ζ), Y( ζ)], Fz( z) P{ Z z}, Dz regio of y-plae such that gy (, ) z, {Z z} {( XY, ) } D z 34.7. Distributio Fuctio fz() z dz P{ z < Z z + dz} fy(, y) ddy D z 34.7.3 Desity Fuctio fz() z dz P{ z < Z z + dz} fy(, y) ddy Eample If the r.v. are idepedet the Eample D z z y Z X Y y z F z f y ddy df z () +, +, z z( ) y fz z fy z yydy (, ), () dz (, ). f () z f () z y f (, y) f ( ) f ( y) ad hece f () z f ( z y) f () y dy f () f ( z ) d y y covolutio of desities. Z X + Y, if z > so the + y z circle with radius z, F ( z) f (, y) ddy, if z <, r / σ z/ σ Fz( z). fy(, y) ( / πσ )ep[ ( + y )/ σ the Fz () z πre dr e, πσ z / σ z > ad fz () z e, z σ z y y z z + y z 999 by CRC Press LLC

Eample 3 r / σ z / σ fy(, y) ( / πσ )ep[ ( + y )/ σ ], Z + X + Y, Fz ( z) πre dr e, πσ z >, fz ( z) ( z/ σ )ep( z / σ, z > Rayleigh distributed, EZ {} σ π/, EZ { } σ, σz ( ( π/ )) σ Eample 4 z If f (, y) f (, y) the F () z f (, y) ddy, f () z yf ( zy, y) dy. The for y y z yz y z ry y ( /[ πσσ r ])ep + ( r ) σ σσ σ y the fz () z of Z X/ Y is fy (, y ) y z rz fz ( z ) [ /( r )] y ep πσσ + dy. ( r ) σ σσ σ But w yep[ y / a ] dy a e dw a ad hece If µ µ the fz () z is Cauchy desity. 34.8 Two Fuctios of Two Radom Variables 34.8. Defiitios f ( z) [( r σσ / π]/[ σ ( z rσ / σ ) + σ ( r )]. z Z gxy (, ), W hxy (, ), D zw regio of the y plae such that gy (, ) z ad hy (, ) w, { Z z, W w} {( X, Y) Dy}, Fzw( z, w) fy(, y) ddy D zw 34.8. Desity Fuctio f zw (z,w) f zw fy(, y) fy(, y) (, z w) + L+ + L, z g( i, yi), w h( i, yi) where ( i, yi) J (, y) J (, y) there are o real solutios for certai values of (z,w) the f (, z w). Jacobia of trasformatio zw are solutios. if Jy (, ) gy (, ) gy (, ) y hy (, ) hy (, ) y 999 by CRC Press LLC

Eample If z a + by, w c + dy the az+ by, y cz+ dw, where a, b, c ad d are fuctios of a,b,c, ad d. Eample a b Jy (, ) ad bc, fzw ( z, w) /[ ad bc ] fy( az + bw, cz + dw) c d z + + y, w / y. If z > the the system has two solutios: zw/ + w, y z/ + w ad, y y for ay w. Jy (, ) / + y y/ + y / y / y ( + w )/( z) ad from 34.8. zw z zw z fzw( z, w) [ z/( + w )] fy, fy,. + w + w + + w + w If z <, f ( z, w). zw 34.8.3 Auiliary Variable If z g(, y) we ca itroduce a auiliary fuctio w or w y. f () z f (, z w) dw. Eample If z y set auiliary fuctio w. The system has solutios y w, y z/ w. J(, y) w ad, hece, z zw fzw(, z w) (/ w) fy( w, z/ w) ad fz( z) ( / w) fy( w, z/ w) dw. 34.8.4 Fuctios of Idepedet r.v.'s If X ad Y are idepedet the Z g( X) ad W h( Y) are idepedet ad sice f zw f f y y (, z w) ( ) ( ) g ( ) h ( y ) 999 by CRC Press LLC

34.9 Epected Value, Momets, ad Characteristic Fuctio of Two Radom Variables 34.9. Epected Value g ( ) Jy (, ) g ( ) h ( ) h ( ) E{( g X, Y)} g(, y) f (, y) ddy; E{ z} zf ( z) dz if z g(, y); E{ g( X, Y)} g(, y ) p, z, PX {, Y y} p 34.9. Coditioal Epected Values discrete case r.v. E{( g X, Y M)} g(, y) f (, y M) ddy; 34.9.3 Momets EgXYX { (, } gyf (, ) ( ydy, ) / f( ) gyf (, ) ( yx dy ) r r µ E{ X Y } y f (, y) ddy, µ R E{ XY} r y µ E{( X µ ) ( Y µ ) } ( µ ) ( y µ ) f (, y) ddy r y r y r y µ σ, µ σ, µ µ 34.9.4 Covariace µ E {( X µ )( Y µ y)} E { XY } µ E { Y } µ y E { X } + µ µ y 34.9.5 Correlatio Coefficiet y y y r E{( X µ )( Y µ )}/ E{( X µ ) E{( Y µ ) } µ / σ σ 999 by CRC Press LLC

34.9.6 Ucorrelated r.v.'s EXY { } EXEY { } { } 34.9.7 Orthogoal r.v.'s EXY { } µ µ µ, µ µ µ, r µ / µ µ 34.9.8 Idepedet r.v.'s f(, y) f ( ) f ( y) Note: y. If X ad Y are idepedet, g(x) ad h(y) are idepedet or EgXhY { ( ) ( )} EgX { ( )} EhY { ( )}. If X ad Y are ucorrelated, the a. b. E{( X µ )( Y µ )}, r y + y + y σ σ σ c. E{( X + Y) } E{ X } + E{ Y } d. EgXhY { ( ) ( )} EgX { ( )} EhY { ( )} i geeral 34.9.9 Joit Characteristic Fuctio Φ ( ω, ω ) E{ e f (, y) e ddy, Ψ ( ω, ω ) l Φ ( ω, ω ) y j ( ω + ω y ) j ( ω + ω y ) y y y j( ω+ ωy) fy(, y) e y( ) d d ( ) Ψ ωω ω ω π Eample Φ ( ω) Ee { } Φ ( ω, ), Φ ( ω) Φ (, ω) jωx y y y Φ ( ω) Ee { } Ee { } Φ ( aω, bω) if Z ax+ by. z jωz j( aωx+ bωy) y if X ad Y are idepedet. Φ ( ω, ω ) Φ ( ω ) Φ ( ω ) y y 999 by CRC Press LLC

34.9. Momet Theorem r Φ(, ) ( + r) j µ r ω ω r 34.9. Series Epasio of Φ( ω, ω ) Φ( ω, ω ) + je{ X} ω + je{ Y} ω { X } ω 4 EY { } ω EXY { } ωω + L+ { } ωω, 4! EX Y + L Ψ( ω, ω ) l Φ( ω, ω ) jµ ω + jµ ω y 34. Mea Square Estimatio of R.V.'s 34.. Mea Square Estimatio of r.v.'s a. a miimizes E{ X a) } if a E{ X} µ b. The fuctio gx ( ) EYX { } regressio curve miimizes y y σ ω rσ σ ω ω σ ω +L E{[ Y g( X)] } [ y g( )] f (, y) ddy r c. y a σ ad b E{} Y ae{} X miimize the m.s. error σ e E{[ Y ( ax + b) ]} [ y ( a b) ] f (, y) ddy e miimum error σ ( r ), r correlatio coefficiet of X ad Y. m 3 y d. If EX { } EY { } the costat a that miimizes the m.s. error e E{( y a) } is such that E{( Y ax) X} (orthogoality priciple) ad the miimum m.s. error is: em E {( Y ax ) Y } a E{ XY}/ E{ X E XY } ad hece em E { Y } { } also EX { }, e E Y E ax m { } {( ) } e E{[ Y E{ Y X}] } m 34. Normal Radom Variables 34.. Joitly Normal If ex { } EY { } the ormal joit desity is: 999 by CRC Press LLC

ry y f(, y) ep +, EX { } σ, EY { } σ πσ σ r ( r ) σ σσ σ 34.. Coditioal Desity rσ f( y) ep y σ ( r ) ( r ) π σ σ, rσ r EYX { } X, E{ Y X} ( r σ σ ) + σ σ 34..3 Mea Value 34..4 Liear Trasformatios EXY { } rσσ, EXY { ] σσ + r σσ E{( X µ )( Y µ )} rσ σ If X ad Y are joitly ormal with zero mea the y Z ax + by, W cx + dy. z y y y σ E{ Z } E{( ax + by) } a σ + b σ + abr σ σ w y y y σ E{ W } c σ + d σ + cdr σ σ, zw z w y y y r σσ E{ ZW} acσ + bdσ + ( ad + bc) r σσ 34. Characteristic Fuctios of Two Normal Radom Variables 34.. Characteristic Fuctio Φ( ω, ω) E{ep[ j( ωx + ωy]} ep[ ( σω + rσσωω + σω )] for E{ X} E{ Y}, ad X ad Y joitly ormal. 34.. Characteristic Fuctio with Meas Φ( ω, ω ) ep[ ( ω µ + ω µ )]ep{ µ ω + µ ω ω + µ ω ], µ meas. j y ij joit momets about the 999 by CRC Press LLC

34.3 Price Theorem for Two R.V s 34.3. Price Theorem If X ad Y are joitly ormal with µ E {( X µ )( Y µ y)} E { XY } E { X } E { Y }, the, a. If µ (r.v. s idepedet) EXY { r } EX { } EY { r } b. c. 34.4 Sequeces of Radom Variables 34.4. Defiitios 3.4.. Defiitios real r.v. X, X, L, X; F(,, L, ) P{ X, L, X } distributio fuctio; f(, L, ) F/, L, desity fuctio. 3.4.. Margial Desities F (, 3) F (,, 3, ) margial distributio for a sequece of four r.v. ; f(, 3 ) f(,,, ) d d margial desity. 3.4..3 Fuctios of r.v.'s E{( g X, Y)} g(, y) f (, y) ddy r µ r r EXY { } r EX { Y } dµ + EX { } EY { } µ EXY { } 4 EXYd { } µ + EX { } EY { } 4 ( µ + EXEY { } { }) dµ + EX { } EY { } µ + 4µ EXEY { } { } + EX { } EY { } 3 4 4 Y g ( X, L, X ), L, Y g ( X, L, X ), µ fy, Ly ( y, L, y) f(,, L, )/ J(, L, ), J(, L, ) g M g L L g g 3.4..4 Coditioal Desities f(, L,, L, ) f(, L,, L, )/ f(, L, ). + + Eample f(, 3) f(,, 3)/ f(, 3), F(, 3) f( ξ,, 3) dξ/ f(, 3) 999 by CRC Press LLC

34.4..5 Chai Rule f(, L, ) f(, L, ) Lf( ) f( ) 34.4..6 Removal Rule f( ) f(, ) d, 3 3 f( ) f(,, ) f(, ) d d, f( ) f(, ) f( ) d 4 3 4 3 4 3 3 3 3 34.4..7 Idepedet r.v. F(, L, ) F( ) LF( ); f(, L, ) f( ) Lf( ) f(, L,,, L, ) f(, L, ) f(, L, ) + + if X, L, X are idepedet of X+, L, X 34.4. Mea, Momets, Characteristic Fuctio 34.4.. Epected Value E{ g( X, L, X )} L g(, L, ) f (, L, ) d Ld 34.4.. Coditioal Epected Values E{ X, L, } f(, L, ) d f(, L, ) d / f(, L, ) a. EEX { { X, L, X}} EX { } b. EXX { X} EEXX { { X, X} EXEX { { X, X} X} c. EX {, L, } EX { } if X is idepedet from the remaiig r.v.'s 34.4..3 Ucorrelated r.v.'s X, L, X are ucorrelated if the covariace of ay two ofthem is zero, EXX { i j} EX { i} EX { j} for i j 34.4..4 Orthogoal r.v.'s EXX { } for ay i j i j 34.4..5 Variace of Ucorrelated r.v.'s σ + L+ σ + L + σ, σz E{ Z E{ Z} } if Z X + jy comple r.v., EZZ { i j} EZ { i} EZ { j} ucorrelated r.v.'s i j, E{ Z Z } orthogoal, are idepedet, 3 3 3 3 i j f(, y,, y ) f(, y ) f(, y ) if Z X + jy ad Z X + jy 999 by CRC Press LLC