This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Stat 116 Homework 6 Due Wednesday, May 21st. Please show work and justify answers. No credit for a final answer with no explanation, even if the answer is correct. 1. Let X and Y be distributed std. Normal with correlation between X and Y equal to ρ . Compute the joint and marginal distributions of X + Y and X − Y , are X + Y and X − Y independent? Solution: Let u = X + Y and v = X − Y so that x = 1 2 ( u + v ) and y = 1 2 ( u − v ). Then the Jacobian is  J  = 1 2 and in the exponent of the bivariate Normal density we have − 1 2(1 − ρ 2 ) ( x 2 − 2 ρxy + y 2 ) = − 1 2(1 − ρ 2 ) ( u 2 (1 − ρ ) + v 2 (1 + ρ )) Therefore the density transforms to f U,V ( u,v ) = 1 4 π radicalbig 1 − ρ 2 exp braceleftbigg − u 2 4(1 + ρ ) − v 2 4(1 − ρ ) bracerightbigg So clearly U and V are independent since there joint density factors into marginal densities, where U ∼ N (0 , 2(1+ ρ )) and V ∼ N (0 , 2(1 − ρ )). 2. Let U ∼ χ 2 ( r ) and V ∼ χ 2 ( s ) be independent for s > 2. (a) Compute E parenleftBig U/r V/s parenrightBig . Solution: Notice U ∼ Γ( α = r/ 2 ,β = 1 / 2) and V ∼ Γ( α = s/ 2 ,β = 1 / 2) so E U = r and E V 1 = integraldisplay ∞ 1 v 2 s/ 2 Γ( s/ 2) v s/ 2 1 e v/ 2 dv = Γ( s 2 2 ) Γ( s/ 2) 2 1 integraldisplay ∞ 2 ( s 2) / 2 Γ( s 2 2 ) v ( s 2) / 2 1 e v/ 2 dv = 2 1 s/ 2 − 1 = 1 s − 2 . 1 So E parenleftBig U/r V/s parenrightBig = s r E U E V 1 = s s 2 , for s > 2....
View
Full
Document
This note was uploaded on 01/13/2010 for the course STATS 116 taught by Professor Staff during the Spring '07 term at Stanford.
 Spring '07
 Staff
 Correlation, Probability

Click to edit the document details