Scribd
Upload
87 views3 pages

APMA 1650: Homework 4

This homework assignment involves 5 problems related to probability distributions and random variables: 1) Find probabilities and expected value related to a given CDF. 2) Use the Poisson distribution to calculate probabilities of defective pixels in TV displays. 3) Calculate variances of Geometric, Pascal, and Poisson distributions. 4) Find variance of a random variable given its expected value and another expectation. Also show a variance inequality. 5) Analyze coin flip data from two students to determine which likely fabricated their results, based on expected numbers of "TT" pairs and "TTT" triples.

Uploaded by

Alexis Lopez
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
87 views3 pages

APMA 1650: Homework 4

This homework assignment involves 5 problems related to probability distributions and random variables: 1) Find probabilities and expected value related to a given CDF. 2) Use the Poisson distribution to calculate probabilities of defective pixels in TV displays. 3) Calculate variances of Geometric, Pascal, and Poisson distributions. 4) Find variance of a random variable given its expected value and another expectation. Also show a variance inequality. 5) Analyze coin flip data from two students to determine which likely fabricated their results, based on expected numbers of "TT" pairs and "TTT" triples.

Uploaded by

Alexis Lopez
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

APMA 1650

Homework 4

Instructions: Homework is due by 11:59pm EST in Gradescope on the day listed on the
course webpage. You can use calculators for this assignment. Solutions must be written
independently and cannot be shared with any other students.

You must show all work and explain your answers thoroughly to get full credit. You will
be graded partly on how well you explain the answers.

There will be a 5pt penalty for homework submitted with problems incorrectly assigned to a
page. A 10pt penalty will be applied for homework submitted during the late window.

1. (18 pts) Consider the following CDF of a random variable X




 0 for x < −2

1/3 for − 2 ≤ x < −1



FX (x) = 1/2 for −1≤x<1 .

5/6 for 1≤x<2





1 for 2≤x

Find the following

a. (4 pts) P (X > −2)

b. (4 pts) P (−1 ≤ X ≤ 2)

c. (4 pts) P (2X ≤ −1)

d. (6 ots) E[X]

2. (20 pts) A TV manufacturer makes a display that has standard 1920 × 1080 pixel array.
Suppose that each pixel independently has a one in a million chance of being defective. Use
the Poisson distribution to answer the following problems. (Hint: you may find the textbook
Poisson CDF calculator very useful).

a. (10 pts) In order for a display to meet the ISO 9241-305 Class II standard, the manu-
facturer can’t have more than 2 defective pixels per one million pixels in that display.
Approximate the probability that a given TV does not meet the Class II standard.
b. (10 pts) Now suppose that the manufacturer needs to make 100 TVs that meet the
Class II standard. Approximate the minimal number of TVs they should manufacture
to be 99% certain they have at least 100 Class II TVs produced. (Hint: make an
educated guess based on the mean and variance of a certain Poisson random variable
and use the CDF calculator to “guess and check” until you find the right value.)

3. (18 pts)
a. (6 pts) Let X ∼ Geometric(p). Calculate Var(X).
b. (6 pts) Let Y ∼ Pascal(m, p). Calculate Var(Y ).
c. (6 pts) Let X ∼ Poisson(λ). Calculate Var(X).
4. (14 pts)
a. (4 pts) Suppose X is a random variable with E[X] = 1 and E[X(X − 2)] = 3. What
is Var(X)?
b. (4 pts) Suppose X is a random variable with mean 1. Show that E[X 2 ] ≥ 1.
c. (6 pts) Let X = IA be the indicator Bernoulli random variable for a probability zero
event A. What is Var(X)?

5. (30 pts) Suppose that for a (very easy) homework assignment students are tasked with
flipping a fair coin 100 times and recording the outcomes. One student does the work and
writes down the results of the 100 flips, another student is lazy and makes up the data. Here
are the outcomes from the two students (in no particular order):

Student 1:
T HHHT HT T T T HT T HT T T HHT HT T HT
HHHT HT HHT HT T HHT T T T HT T T HT H
T T HHT T T T T T T T HT HHHHHT HT HT H
T HT HT HHHHHT HHT T T T T HT T HHT H
Student 2:
HT T HT T HT HHT T HT HT HT T HHT HT T
HT T HHHT T HT T HT HT HT HHT T HT T H
T HT HT HT HHHT T HT HT HT HHT HT T T
HT HHT HT HT HT HHT T HT HT HT T HHT
Your goal is to figure out which student was most likely to have fudged the data. Your
strategy is to count the number of T T pairs in each sample. (Here counting T T pairs
involves counting how many times T T shows up out of all 99 neighboring pairs. For instance
for the 8 tosses T T T HT HT T has 3 T T pairs.)
a. (12 pts) Suppose you flip a fair coin n times. What is the expected number of T T pairs?
(Hint: Use sums of Bernoulli random variables to count the number of occurrences.)

b. (3 pts) Based on your answers to a, which student do you think most likely fudged
their data.

c. (15 pts) Repeat a and b for the number of T T T triples (i.e. T T T T T has 3 triples).

You might also like