EECS 126 Midterm 2: Probability and Statistics, Exams of Probability and Statistics

A midterm exam for the university of california, berkeley, eecs 126 course in probability and statistics. It includes questions on joint probability density functions, system reliability, and estimating population mean. Questions cover topics such as finding probabilities, expectations, and minimum mean square error estimators.

Typology: Exams

2012/2013

Uploaded on 03/22/2013

farhan
farhan 🇮🇳

3

(1)

55 documents

1 / 2

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Name: _________________________ Student ID No: ______________
UNIVERSITY OF CALIFORNIA
College of Engineering
Department of Electrical Engineering and
Computer Sciences
Professor Ren Fall 1997
EECS 126 — MIDTERM #2
November 17, 1997, Monday 7-9 p.m.
[42 pts.] 1.
Given the joint probability density of two RVs and
a)
Find the value of , and the cdf . (6 pts.)
b)
Find
. (6 pts.)
c)
Find the probability that . (6 pts.)
d)
Find
. (6 pts.)
e)
Find the minimum mean square error estimator of given . Compute the resulting
mean square error. (6 pts.)
f)
Find the linear minimum mean square error estimator of given . Compute the
resulting mean square error. (6 pts.)
g)
Are and independent? Uncorrelated? Orthogonal? Explain your answer. (6 pts.)
[35 pts.] 2.
An electronic system has components. Let the lifetime of each component be
, in hours. Assume that , are mutually independent,
and have identical density . Let the lifetime of the system be .
a)
Suppose the system works only if all components work. Find the pdf and expectation
of . (10 pts.)
b)
Suppose we already know that the system has already lasted 10 hours. Find the condi-
tional pdf and expectation of . (12 pts.)
c)
To increase reliability, we use redundancy by increasing the number of components from
to . Suppose the system works so long as there are at least components working.
Find the cdf of . (13 pts.)
X
Y
fXY x y,( ) k x y+( ) 0x1 0 y1 ,
0 otherwise
=
k
FXY x y,( )
FXx( ) FYy( ) fXx( ) fYy( ),,,
X Y 1 2
X
Y
X
Y
X
Y
n
Xi i,1 2 n, , ,=
Xi i,1 2 n, , ,=
fXix( ) ex x0,=
Y
n
Y
Y
n
2n
n
Y
pf2

Partial preview of the text

Download EECS 126 Midterm 2: Probability and Statistics and more Exams Probability and Statistics in PDF only on Docsity!

Name: _________________________ Student ID No: ______________

UNIVERSITY OF CALIFORNIA College of Engineering Department of Electrical Engineering and Computer Sciences

Professor Ren Fall 1997

EECS 126 — MIDTERM

November 17, 1997, Monday 7-9 p.m.

[42 pts.] 1. Given the joint probability density of two RVs and

a) Find the value of , and the cdf. (6 pts.) b) Find. (6 pts.) c) Find the probability that. (6 pts.) d) Find. (6 pts.) e) Find the minimum mean square error estimator of given. Compute the resulting mean square error. (6 pts.) f) Find the linear minimum mean square error estimator of given. Compute the resulting mean square error. (6 pts.) g) Are and independent? Uncorrelated? Orthogonal? Explain your answer. (6 pts.)

[35 pts.] 2. An electronic system has components. Let the lifetime of each component be , in hours. Assume that , are mutually independent, and have identical density. Let the lifetime of the system be.

a) Suppose the system works only if all components work. Find the pdf and expectation of. (10 pts.) b) Suppose we already know that the system has already lasted 10 hours. Find the condi- tional pdf and expectation of. (12 pts.) c) To increase reliability, we use redundancy by increasing the number of components from to. Suppose the system works so long as there are at least components working. Find the cdf of. (13 pts.)

X Y

f (^) XY ( x y , ) k x (^^ + y )^0 ≤^ x^ ≤1 0,^ ≤^ y^ ≤^1 ^0 otherwise

k FXY ( x y , ) FX ( x ) , FY ( y ) , f (^) X ( x ) , f (^) Y ( y ) XY ≤ 1 ⁄ 2 f (^) X Y ( x y ) X Y

X Y

X Y

n Xi , i = 1 2, , … , n Xi , i = 1 2, , … , n f (^) Xi ( x ) = e –^ x , x ≥ 0 Y

n Y

Y

n 2 n n Y

2 of 2

[23 pts.] 3. Let be a sequence of i.i.d. RVs with mean and unit variance. Suppose is

unknown. a) Propose a scheme to estimate from. (5 pts.)

b) Suppose your estimate of based on is denoted as. Using Central Limit Theorem, find a range of that would guarantee the quality of the estimate in the follow- ing sense:

. (18 pts.)

X 1 , X 2 , … μ μ

μ X 1 , … , Xn

μ X 1 , … , Xn μˆ (^) n n

P ( μˆ^ n – μ ≤0.1) ≥0.