{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

mtsample-sol

# mtsample-sol - EE 278 Statistical Signal Processing Sample...

This preview shows pages 1–2. Sign up to view the full content.

EE 278 Saturday, July 18, 2009 Statistical Signal Processing Handout #9 Sample Midterm Solutions 1. (20 points) Inequalities. a. (2 points) None . Since P( A | B ) = P( B | A )P( A ) P ( B ) and P( A ) can be larger or less than P( B ), no equality or inequality holds in general. b. (2 points) . Since p X ( x ) = Σ y p X,Y ( x, y ) and p X,Y 0, p X ( x ) p X,Y ( x, y ). c. (2 points) None . Here f X ( x ) = integraltext −∞ f X,Y ( x, y ) dy , so we cannot in general say how f X ( x ) relates to f X,Y ( x, y ). If f X ( x ) > 0, then f Y | X ( y | x ) = f X,Y ( x, y ) /f X ( x ) can be greater or less than 1. d. (3 points) =. Since expectation is linear, E parenleftbigg X 1 X 1 + X 2 parenrightbigg + E parenleftbigg X 2 X 1 + X 2 parenrightbigg = E parenleftbigg X 1 + X 2 X 1 + X 2 parenrightbigg = E(1) = 1 . e. (3 points) . Let g ( x ) = x log 2 x . Then g ( x ) is convex since g ′′ ( x ) = 1 / ( x ln 2) > 0 for x > 0. Now, by Jensen’s inequality and E ( X ) = 2, E( X log 2 X ) E( X ) log 2 E( X )) = 2 log 2 2 = 2 . f. (4 points) . Consider P { ( XY ) 2 > 16 } ≤ E (( XY ) 2 ) 16 radicalbig E ( X 4 ) E ( Y 4 ) 16 = 1 8 , where the first inequality follows from the Markov inequality and the second inequality follows from the Schwarz inequality.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}