——____Unit-0a Broriole Random Variables _ #4 a - - | 4 and_y_be two rondom_uasiatle's defined on the sample jpoce ‘5’, Then the gandom_teclor x. Y-assigns a poinl inthe toro dimensional space *R?? is. called Tivo dimensional. sandom vasable ot | Examples : Heigl's and weight’s of group of person's. dncome_and Expenditure of group_of family. Volume and Pre sure of agar. - Marginol’s 3. Conditional. condom Variables seperatel #|| Bivoriote Distribution: i = | Bivariate disteibution canbe debined fos two kinds ep tordein Variable ie. Diserele ond Continuous =: There aye three Concepts , Bo1_both Discrete and _ B09 a9ay ee x || Buoriate Discrete Random Uasiables: i. The _joint_, morginal_and_condiifional's for cliscsele sand Variables con be explained below, [Link] Psobability mass function : het x and y be too discrete sandom _uvariable's foking the Uolues 21, X2,.--. Xm and Yr Yas=--Yn respectively » The _furetion ij defined on x: Xi, ysy; is Caled Joint probability marr _ Gucction of 214 and is denoted by plx: xi, Yeyi Jog P(xiny;* and 5 2epresented by the following Bivasiale Table. _yo We pu 7 yp | pat - i wen! pin — . Yf. pos Masginols of | pt. sobability mon function ¥ and. ishy the bollovring Conds Plexi, Yr yj Jand it Should ste bers, |r eGiyj deo F fi plaisyidet— tnaxgial a A function « Of ies ) Marginal probobitity function of yo || Marginal Probability _ ___Marginal_ Probe | Marginal Probability Bunciron of y Pp PYe yy.Sj Py. 4 3-Conditional Pacbability Function: P(yeyjd: PC nex, + Pl a oliowsing _¢ Ply:yjp%o oj He peje 2 eCyeyres = co There are two types of Conditional Diohability function's onal Probability urction ef y given x i) Conditional Probability gurctien of x Yiven_y. 1 Conditional Probability function ef y gixen x: _— The Conditional Probability Bunction of *y * given x7 Plysyj/ x: xi} os Ply/s), s_cenoted by _is_deGined av PLYsy/ xem]: plysxi.y: Pixs xi] |__Paoperties of Conditional probability function ey ) Ply: 43 /x=aiJ 20 Y given x 2 £ P Lye yj /x: oi] 4 ii) Conditional Prob. ty function of xX given Y - The Conditional Probability plxea/y: yj} _o Pp l/y]. lution of ‘x’ given'y".s dented by fis defined os plxewi ly: gj]: pfx: xi yey] Plysy) Paopertics of conditional Probability function of X given y I Plxz ai/y2 gi Jzo0 2) F plxsai/y: yj de 2\ i | i 03-09 a0aG, x Wa. dont Probabitily clensity Lanclion : om. Uasiable + She joint probaly. deny rn veeriatle ¥en 15. enoted by, Bivariate. Distaibulion , Con tinuous Ron funtion of the randor fay lay) or Gla y) 08 Pay (Xe > yry ib should. Stoticgy the following Condilions ,——— 3) POuy ro 0 LEP bGayrdedys 2 ginal Probability. density _gunetion 2 The _rmarginal probability clensity. function ob X. | far os tar ond_is_deffined_by frlx0= £° Play dy + Phopestits_of Marginal _probabilily density Bunction_of_x; by byLy) or BtG)_and is dened by by (yrz $° Play )dz. + Properties of Marginal probability density fucetion of ¥; J hizo _ nl oS” £Gyddyet |3. Conditional Probability density function: Thee area —Seodilional probobility density Gunction, | 4 Constonal_pbobitiy_demity function. of Y givin x + Conditional ili Pi Conditional Probability density furctron of * given] => Conditional Probability cfensit jiven Xi Centitenat Psbabilty denuty function 8 given. is_ dete 14 Ly/x (yle) or as ~ pg vi0- a ai 01 Ee an sien by. lad: fig? fi —_ _ ————— | aan a = _ |e —}~—— ry is denoted by EI 2) 0 £0) ord i8 giver by CONy d= — Fiche bey) 20 ‘Y, ———f}a-fe4rr0 | ‘Sb J ; ; / ie ene OE byldedy sa | Conkitional Probability domly function of x given * Properties: WD fybd > >, Conditional Proberbitity cnnity function of x gen | -*_Paoperties [2S 8 Gxly)-des 4 + || Distribution \ Bunction of 12 doint Distribution Function : ave real numbers then, Pars x$ bi; 02< vsb2)# Fxy (bs ,b2 4 Fxy(or.a2)-Fxy (ar, b2 =F xy(bi,a2) -id_Limits of __joint_distribution function i$ 0 and1. ie 0 _F¢ ii F(-0 F(a. 0 ECoy)=0 £ (2.70) 20 ~ iv) 26 £ (24) 16 the jolnl probability: clesily function of x,y then lay) 22°F (uy) ox. ay__The_Masginal_distubution bunction of X48 given by, 15_denoted by FX (x ixsxJ 2. Marginal Distribution Function _Fx (a) and it fribution function of x; Properti |i) P(acx 4b): F(b)- F(a). es : ii) OS Fyxexys 1 - —— | iii) F-@ 20, F (ode 4 ———— Similarly The Marginal distribution Bunction of Ys given ty, Eyly) ond it is denoted by FyCy)= PLYS 4] - = Pl xem, YSy) = eplxeme¥ey] ssn 2 F Coy) ¥ Paopesties ‘of Marginal distribution function of Y; i _P (Csysd) = F(d)- FCc) m_OS Fy(y)s2 in) _F(-w)=0, F(t): 4 3. Conditional_Distvibution Function + ee it) The Conditional distribution function of X given Y. 3) The Conditional _ distribution function op _ _ Conditional Distribution Function of Y given x: —— The Conditional distribution function of y given x_is_givtt- Ir ie Fyfe Cyne Fyfe) = Ply sy [xex] + Properties: n eles ysd/xJ=& (d/e = ECe/y7 lpose(y/xds23)_FCo/x)=0, F (oo/x) 24 Conditional Distribution Function of x _given-y: The Corditional Distribution Function of x given Yi ven Fly (xfy) = €Cxly) z PL xx /¥sy] + Properties: QP [ac xs biyd= e(b/y)-F (aly) 2) OS FCx/y) S14 | 5) E (-w/y)z0, F(wAy)=4 %1|_ Endependent_ Variables : Definition: Tivo Random Variables 4 and Y ane snd to be independent Band only if P (x.y): P(x) Ply? fos discrete ard $lxay= £0 £Y) for contincous... ‘This defination is also poo for distribution Bunctron that i: Two _tardom variable x and y: axe 4aid to be independent if. and only ip Flag): FC. 6g) whee Eley) 6 joint distributi -0n function F(x) and Fly) aye marginal distsibution Bunctions of x and y. - X||_Co-Variance using Expectation: Gf x and Y are to tandom variables then co-variance _ between the variable's x andy «$ defined a3 Cor Gay) £ (Cx) y-£49)) This_formula can also be evpresed as, €(xy-xety)- VEC EGY) | 2. € (xy) EC ECy) = EQ r(x) + ECO ECYT = ECxy)- ek? Ely) || Gndependence of Redom Varinbles Using Expectation: Gf __sandom variable x ard y are said to be independent if E (xy) = E(x). Ey) ie Cov (x.y) 20y || Addition Theorem _of Expectation; Stotement: dp x and y_are_fwo_rondom variables then, E(xty)= EGO+ EY) 7 Proop: lef _x and y be ao continuous tandom variables. cais, joinl_pab, £(x.4) and_marginel pd g's. £errand £6) recy, (we can prove this theorem for discrete Yandom Gariables by — using £€? instead op ‘5? ), —— Ky __the_defini-tion_af expectation, EG: SE x Bleddx eq: S ye ydy f, Nooo consider € Cxsy d= Ie or (x+y 9B be ro 2 Se IS Rey rdedy + JOSS y Eleyddady i 7 2 SP x (SP £ faery dy Jd t $° gC fe {lus * ke Mow Bsom the definition off marginal distributron Buncsion ¥= Blz) and #2 Pty) Theabove equation becomes, § xhtrdx+ (° yPlyrdy a EGuy)= Cx) + Ey) Multiplication Theorem of Expectation: Statement: @_x and y are independent _sandom variables He, WE (xy)= Ex- ey ¥ ee Prog: Leb x ond y be inclep endent continuous random varias then by definition 0€ expectation - EOD: SJ" xPardx| coher 862 and Ely) are ihe marginal pdf | Proog: let x and y be ts0 cont | bO & bey) _E (axs by): “of x and y (ue Now _E (xy) = fe ey foe dxdy _ there 66004) 35 the joint probability density function « Elo OL sey fo pup dxdy (x ey ov indepeodent £ (1.4): fon ty). : Peodx So 4 Purdy Theetore- by the definition, Eby) = E60. Ely) Prats oa abl D4 bey) aa UOUA Fandom variables then, ve _the marginal pdf's of x and y then, =e oe —— the joint probability o ely function cf x and y cobere (xy? i —£ (axrby) = a ne ax foe dxdy + City Eloyldxdy = E (ax+ by) ——_}-_____ | Now _fsom__the |x + fe0_and oa = : Bo WE Coxsbyts oe xlfedes bat y pyidy 2 E Caxsby): o£ I+ bE) 7 19-09-2080, ee . >| vCaxs oy) =o? v(x >t b? vey 2ab Cov (x.y ) Prope: v(axsby): € (oxs by - Elax+by))? u(axs by) + £ (axtby - aE Cx?- bey)? wted: ECE-E v(axsby) = E(alx-€x ds b (y- E(y)))* ae = e vaxs by) = aE x- Ex)? 4 bey E(u)’ # 2abE (x= £0 )(Y- EK) ¥ Coxrby) = a2 vx) + b? vey)+ 20b CovCxry)Ey Co- Variance: . Covariance of x and y iv given by, cour) £ (CMM) Properties of CoMlarianc! sng ss _ Shows that Coulax. by): ab Cov (xy) 9e_Kro%o that Cov Cax,by) +E ((ax-€ (and) (by -e(by))) Cov Can, oy)=E((ax- of) Cby bE) a te nts Cov ifan tye E(a Ce eed beye ety)) Cou (ax. byy? ab E(O-exn(y-eyy) Gu_(ax, by): ab Covlxy) |Show that Cov (xta,yrb): Covey) Prov Cou [Link])= E(CHsa-€ (x10))(ytb - E(ysb)) Cov Cato, yr BIZ (Cx4e Ebdon) (yt b= ECI-b) Cov (x10; y rb) CKRp EOI) (yey) — Cou (xta, yro)*' Cov (x.y) 2 | Shove thot Cou lar-tb,.cytd = acConCxy) — aa |-Proop: Cov Carib, cyid): € (Cart -eCaxrory(cys d= E (cysd)) —— —Cov Coxe, cy td): € (@ x4 ECB) (cy +d - CE Cy) -d) —— ——Cou_Coxib, cy td): € Ca (x1 Go)c ly ow | _Cov_(axib, cyed)= cE ((x-€¢x Ny. Cov Cox, cyrd):0 _