Unit 03 - Random Variables - 1 Per Page
Unit 03 - Random Variables - 1 Per Page
Unit 03 - Random Variables - 1 Per Page
S = {TTT,HTT,THT,TTH,THH,HTH,HHT,HHH}
3
Random Variable as a function
(mapping)
4
Definition: Distribution
• Let X be a random variable. The distribution of X is the
collection of all probabilities of the form P(X ∈ C) for all
sets C of real numbers such that {X ∈ C} is an event.
• What the heck does that mean?
• Really, it’s just listing/defining what all of the probabilities
are for all of the possible values of X (of course, the values
of X are on the real line).
• The definition is written in specific language to eventually
allow for continuous random variables (that is, random
variables that can fall anywhere within a specific range on
the real line. Example: time it takes to complete an exam).
5
Distribution of a Random Variable
• Be careful: you can only find probabilities of events. So be
sure to use proper notation.
• P(X = 2) is good notation. P(X) or P(2) is bad.
6
Unit 3 Outline
• Random Variables (RVs) and Distributions
• Discrete RVs and Probability Mass Functions (PMFs)
• Bernoulli and Binomial Distributions
• Hypergeometric Distribution
• Discrete Uniform Distribution
• Cumulative Distribution Functions (CDFs)
• Functions of Random Variables
• Independence of Random Variables
7
Discrete Random Variables
8
Probability Mass Function (PMF)
• For a discrete random variable X, the probability mass
function (or simply just probability function) of X is
defined as the function f such that for every real number x,
f ( x) P( X x)
• The closure of the set {x: f(x) > 0} is called the support of
the distribution of X.
• Examples:
• X = # heads in 3 flips of a coin.
• T = sum of two fair 6-sided dice
9
Plot of a discrete PMF
• T = sum of two fair 6-sided dice
10
Unit 3 Outline
• Random Variables (RVs) and Distributions
• Discrete RVs and Probability Mass Functions (PMFs)
• Bernoulli and Binomial Distributions
• Hypergeometric Distribution
• Discrete Uniform Distribution
• Cumulative Distribution Functions (CDFs)
• Functions of Random Variables
• Independence of Random Variables
11
Story of the Bernoulli Distribution
• Imagine you are going to flip a coin once, with known
probability p of showing heads.
• Let X = # heads.
• What is X’s distribution? That is, what is the probability mass
function for X? (don’t forget to mention X’s support)
• Extension #1: you are going to perform an experiment once
that has probability p of being a success . Let X = # successes.
• Extension #2: you are going to randomly sample a single
individual from a population and measure whether or not they
have a specific characteristic, and p is the proportion of
individuals in the population with that characteristic.
• Example: did one randomly sampled Massachusetts voter
vote for Obama in 2012 (p = 0.60)?
12
Bernoulli Distribution Definition
• An r.v. X is said to have the Bernoulli distribution with
parameter p if P(X = 1) = p and P(X = 0) = 1 – p, where
0 < p < 1. We write this as X ~ Bern(p). The symbol ~ is
read “is distributed as".
• Any r.v. whose outcomes are just 0 and 1 has a Bern(p)
distribution, where p represents the probability of X = 1.
• The number p is called the parameter of the distribution; it
determines which specific Bernoulli distribution that X
follows.
• This means that there is not just one distribution, but rather
a family of Bernoulli distributions, each defined by this
parameter p.
13
Indicator r.v. and Bernoulli Trials
• The indicator random variable of an event A is the r.v.
which equals 1 if A occurs and 0 otherwise. We will denote
the indicator r.v. of A by IA or I(A). Note that IA ~ Bern(p)
with p = P(A).
• An experiment that can result in either a “success“ or a
“failure" (but not both) is called a Bernoulli trial. A
Bernoulli random variable can be thought of as the
indicator of success in a Bernoulli trial: it equals 1 if
success occurs and 0 if failure occurs in the trial.
14
Story of the Binomial Distribution
• Imagine you are going to flip a coin n times, with known
probability p of showing heads each times. Assume the result
of each flip is independent from one another.
• Let X = # heads.
• What is X’s distribution? That is, what is the probability mass
function for X? (don’t forget to mention X’s support)
• Imagine instead you are repeatedly performing an experiment
n times, where the probability of success, p, for each trial is
fixed and each trial is independent from one another.
• Let X = # successes.
• Imagine instead you are sampling n individuals from a huge
population (so large that it can be assumed to be infinite), and
are measuring each individual for a specific characteristic.
15
Binomial Distribution Definition
• Suppose that n independent Bernoulli trials are performed,
each with the same success probability p. Let X be the
number of successes. The distribution of X is called the
Binomial distribution with parameters n and p. We write
X ~ Bin(n, p) to mean that X has the Binomial distribution
with parameters n and p, where n is a positive integer and
0 < p < 1.
• If X ~ Bin(n, p), then the PMF of X is:
n k
P ( X k ) p (1 p ) n k
k
for k = 0, 1, …, n.
16
Plot of
Binomial
PMFs
17
The Binomial Distribution: Example
• If a couple are both carriers of a certain disease their child
has probability 0.25 of being born with the disease.
Suppose that a couple has 5 children:
• What is the probability that none of their children have
the disease?
• What is the probability that at least two children have
the disease?
• Let’s use the formula to calculate the answers, and check
online:
• Binomial Calculator online:
http://www.stat.tamu.edu/~west/applets/binomialdemo.html
18
Binomial Distribution: More Examples
• Let X ~ Bin(n = 20, p = 0.3).
• What is P(X = 10)?
• What is P(X ≥ 10)?
• What is P(X > 10)?
• What is P(X < 20)?
• 20% of Harvard upperclassmen are varsity athletes. A
random sample of n = 64 students were selected from the
Quad houses, and only 8 of them were varsity athletes. [If
Harvard housing was assigned randomly], what is the
probability of selecting 8 or fewer athletes in this sample of
64 students?
• When is a Binomial distribution symmetric?
• Let Y = n – X. What distribution does Y follow?
19
Unit 3 Outline
• Random Variables (RVs) and Distributions
• Discrete RVs and Probability Mass Functions (PMFs)
• Bernoulli and Binomial Distributions
• Hypergeometric Distribution
• Discrete Uniform Distribution
• Cumulative Distribution Functions (CDFs)
• Functions of Random Variables
• Independence of Random Variables
20
Story of the Hypergeometric Distribution
• Suppose we have an urn filled with w white and b black balls.
Then drawing n balls out of the urn with replacement yields a
Bin(n, p = w/(w+b)) distribution for the number of white balls
obtained in n trials, since the draws are independent Bernoulli
trials, each with probability p = w/(w+b) of success.
• If we instead sample without replacement, as illustrated
below, then X = the number of white balls follows a
Hypergeometric distribution.
• What is the PMF for X? (be sure to mention the support)
21
Hypergeometric Distribution Definition
• Consider an urn filled with w white and b black balls. We draw n
balls out of the urn at random without replacement, such that all
w+b
samples are equally likely. Let X be the number of white
n
balls in the sample. Then X is said to have the Hypergeometric
distribution with parameters w, b, and n; we denote this by X ~
HGeom(w, b, n).
• If X ~ HGeom(w, b, n), then the PMF of X is:
w b
k n k
P( X k )
w b
n
for integers k satisfying 0 ≤ k ≤ w and 0 ≤ n – k ≤ b.
22
Hypergeometric Distribution Examples
• In a five-card hand drawn at random from a well-shuffled
standard deck, the number of aces in the hand has a
HGeom(4, 48, 5) distribution, which can be seen by thinking
of the aces as white balls and the non-aces as black balls.
• What is the probability that the hand has exactly 3 aces?
• A forest has N deer. Today, m of the deer are captured, tagged,
and released into the wild. At a later date, n deer will be
recaptured at random. Assume that the recaptured deer are
equally likely to be any set of n of the deer, e.g., a deer that
has been captured does not learn how to avoid being captured
again. Let X = # deer in the n recaptured deer that were
captured the first time.
• What distribution does X have?
23
Hypergeometric vs. Binomial
• The Binomial and Hypergeometric distributions are often
confused. Both are discrete distributions taking on integer
values between 0 and n for some n, and both can be
interpreted as the number of successes in n Bernoulli trials
(for the Hypergeometric, each tagged deer in the recaptured
sample can be considered a success and each untagged deer a
failure).
• However, a crucial part of the Binomial story is that the
Bernoulli trials involved are independent. The Bernoulli trials
in the Hypergeometric story are dependent, since the sampling
is done without replacement: knowing that one deer in our
sample is tagged decreases the probability that the second
deer will also be tagged.
24
The HGeome is connected to the Binome
• The Binomial and Hypergeometric distributions are
related in a couple very specific ways:
25
Unit 3 Outline
• Random Variables (RVs) and Distributions
• Discrete RVs and Probability Mass Functions (PMFs)
• Bernoulli and Binomial Distributions
• Hypergeometric Distribution
• Discrete Uniform Distribution
• Cumulative Distribution Functions (CDFs)
• Functions of Random Variables
• Independence of Random Variables
26
Discrete Uniform Distribution
• Let C be a finite, nonempty set of numbers. Choose one of
these numbers uniformly at random (i.e., all values in C are
equally likely). Call the chosen number X. Then X is said to
have the Discrete Uniform distribution with parameter C;
we will denote this as X ~ DUnif(C).
• The PMF of X ~ DUnif(C) is:
1
P( X x)
C
for x ∈ C (and 0 otherwise).
• Common example: random integer from a to b.
27
Discrete Uniform Dist. Example
• There are 100 slips of paper in a hat, each of which has one of the
numbers 1,2,…,100 written on it, with no number appearing more
than once. Five of the slips are drawn, one at a time.
• Consider random sampling with replacement (with equal prob’s):
a) What is the distribution of how many of the drawn slips have a
value of at least 80 written on them?
b) What is the distribution of the value of the jth draw (for 1 ≤ j ≤ 5)?
c) What is the probability that the number 100 is drawn at least once?
• Now consider random sampling without replacement (with all sets
of five slips equally likely to be chosen):
d) What is the distribution of how many of the drawn slips have a
value of at least 80 written on them?
e) What is the distribution of the value of the jth draw (for 1 ≤ j ≤ 5)?
f) What is the probability that the number 100 is drawn at least once?
28
Unit 3 Outline
• Random Variables (RVs) and Distributions
• Discrete RVs and Probability Mass Functions (PMFs)
• Bernoulli and Binomial Distributions
• Hypergeometric Distribution
• Discrete Uniform Distribution
• Cumulative Distribution Functions (CDFs)
• Functions of Random Variables
• Independence of Random Variables
29
Cumulative Distribution Function (CDF)
• The cumulative distribution function of a r.v. X is the
function FX given by FX(x) = P(X ≤ x). It is often written as
just capital F without the subscript, or F(x). (other letters,
like G or H, can also be used). (DeGroot p.108)
• Properties of CDFs:
• CDFs are always non-decreasing
• limx→-∞[F(x)] = ___. limx→∞[F(x)] = ___.
• CDFs are always continuous from the (right or left?)
• Example: Let X ~ Bin(n = 4, p = 0.5).
• Find F(1.5).
• Find F(x) for all x.
30
Plot of a CDF
• Let X ~ Bin(n = 4, p = 0.5).
31
PMFs vs. CDFs
• We have seen three equivalent ways of expressing the
distribution of a random variable. Two of these are the PMF
and the CDF: these two functions contain the same
information, and we can always figure out the CDF from the
PMF and vice versa. Generally the PMF is easier to work
with for discrete r.v.s (the CDF requires a summation).
• A third way to describe a distribution is with a story that
explains (in a precise way) how the distribution can arise. We
used the stories of the Binomial and Hypergeometric
distributions to derive the corresponding PMFs. Thus the
story and the PMF also contain the same information. We can
often achieve more intuitive proofs with the story than with
PMF calculations, algebra, etc...
32
Unit 3 Outline
• Random Variables (RVs) and Distributions
• Discrete RVs and Probability Mass Functions (PMFs)
• Bernoulli and Binomial Distributions
• Hypergeometric Distribution
• Discrete Uniform Distribution
• Cumulative Distribution Functions (CDFs)
• Functions of Random Variables
• Independence of Random Variables
33
Function of a R.V.
34
Illustration of a function of a r.v.
35
Function of a R.V. Example
(a Random Walk)
• A particle moves n steps on a number line. The particle starts
at 0, and at each step it moves 1 unit to the right or to the
left, with equal probabilities. Assume all steps are
independent. Let Y be the particle's position after n steps.
Find the PMF of Y.
• Hint: how does each step relate to a Bernoulli trial?
• Let X ~ Bin(n, p = 1/2). Then Y = 2X – n → X = (Y + n)/2.
• Thus:
1
n
n
P (Y k ) P ( 2 X n k ) P ( X ( k n) / 2)
( k n) / 2 2
• For k = -n, -n+2, -n+4…,n.
• Sanity check: does that answer make sense?
36
PMF of a function of a discrete R.V.
P( g ( X ) y ) P( X x)
x: g ( x ) y
38
Be careful!
• A common error in probability is to confuse a random variable
with its distribution. Two completely different random variables
could very well have the same distribution.
• Other common errors:
• Given a r.v. X, trying to find the PMF of 2X by multiplying the
PMF of X by 2. Think about what this means for the plot of X
and it’s distribution.
• Claiming that because X and Y have the same distribution, X
must always equal Y (i.e. P(X = Y) = 1). Consider flipping a
fair coin once. Let X = indicator it lands heads while Y = 1–X
indicator it lands tails. Both X and Y ~ Bern(0.5), but X = Y is
impossible. The PMFs of X and Y are the same function, but X
and Y are different mappings from S → ℝ.
• Always ask yourself: does the result make sense???
39
Unit 3 Outline
• Random Variables (RVs) and Distributions
• Discrete RVs and Probability Mass Functions (PMFs)
• Bernoulli and Binomial Distributions
• Hypergeometric Distribution
• Discrete Uniform Distribution
• Cumulative Distribution Functions (CDFs)
• Functions of Random Variables
• Independence of Random Variables
40
Independence of two R.V.s
P ( X x, Y y ) P ( X x) P (Y y )
for all x, y ∈ ℝ.
P ( X x, Y y ) P ( X x) P (Y y )
for all x in the support of X and y in the support of Y.
41
Independence of many R.V.s
P ( X 1 x1 ,..., X n xn ) P ( X 1 x1 )...P ( X n xn )
for all x1,…, xn ∈ ℝ. For infinitely many r.v.s, we say that
they are independent if every finite subset of the r.v.s is
independent.
42
Independence of R.V.s example
• Suppose you roll two dice, with X the number on the first
dice and Y the number on the second dice.
• Let U = X + Y and V = X – Y.
43
Conditional Distributions and PMFs
• Just like with events, we can equivalently show that two
r.v.s are independent based on conditional probability.
• We need to define the conditional distribution:
• For any discrete r.v.s X and Z, the function P(X = x| Z = z),
when considered as a function of x for a fixed z, is called the
conditional PMF of X given Z = z.
• Thus, two discrete r.v.s, X and Z, are independent if
P(X = x | Z = z) = P(X = x) for all possible values of x and z.
• We’ll come back to this idea of conditional distributions
later.
44
Independent and Identically Distributed
R.V.s (i.i.d.)
• We will often work with random variables that are
independent and all came from the same distribution. These
r.v.s are called independent and identically distributed, or
i.i.d. for short.
• What is the difference between independent and identically
distributed?
• Give examples of:
• independent and identically distributed r.v.s
• independent and not identically distributed r.v.s
• dependent and identically distributed r.v.s
• dependent and not identically distributed r.v.s
45
Binomial is a sum of i.i.d. Bernoullis
46
Last Word: what is random?
47