Question

6. Suppose that (X1, X2) ~ N(0,0,1,1,0). It follows from this that the joint PDF of (X1, X2) is given by 1 fx1,x2 (x1, x2) - cxp («** + x2)) Furthermore, if 1 and (X1 + X2) ✓2 1 Y2 (X1 - X2) V2 Then (Yı, Y2) ~ N(0,0,1,1,0) as well. (a) If X1 < X2, what are the possible values of Y¡ and Y2? (b) If Y2 < 0, what are the possible values of X1 and X2? (c) What is the marginal distribution of Y2? (d) Find P(X1 < X2) (be sure to justify your answer).

Answer 1

The detailed solution is given in the pictures below.

Please go through them carefully specially the notations.

Hope the solution helps. Thank you.

(Please do comment if further help is required)

vid Also xat Xz but can take clearly X, X2 N(0,1). [As correlation of zero for bivariate į lid Neo, 1) normal inplies independence] fa x, ex > < 2x2 ER and X - X2 <o Thias Y can take real values any negative real rabel. That is, Ye and Y. € (2000) The set of possible value for (Y,,Y) is sa { Cherya): y, en and yet to)}. Ang x, exz set of possible values that (X, X₂) take is { (22) : x, Elh and -oo <n, <m₂ < 00} (0) abready got that (Arg) That is Y MCO, 1) [Property of bivariate Ans: Marginal distribution of is No1). normal distribution] (d) P(x, y) = P(x - x 20) =P(Y 20) = P(N(0,1)<o), o is the mediat of Yzi Ans: 0.5 com t. X - Xzco can we id No!) 7 as 4 No,