Assume each is fair Random Sample n 5 w or wo replacement Random Variable X Spades in sample x 0 1 2 3 4 5 Outcomes of X 2 0 0 0 1 1 0 0 1 0 1 ID: 760292
Download Presentation The PPT/PDF document "Infinite Population of decks of 52 card..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Infinite
Population
of decks of 52 cards
(Assume each is fair)
Random Sample
n
= 5
(w/ or w/o replacement)
Random Variable
X =
# Spades in sample
(
x
= 0, 1, 2, 3, 4, 5)
Outcomes of “X = 2”
(0, 0, 0, 1, 1)
(0, 0, 1, 0, 1)
(0, 0, 1, 1, 0)
(0, 1, 0, 0, 1)
(0, 1, 0, 1, 0)
(0, 1, 1, 0, 0)(1, 0, 0, 0, 1)(1, 0, 0, 1, 0)(1, 0, 1, 0, 0)(1, 1, 0, 0, 0)
Hypergeometric Distribution
Binary
r.v
.
10 combinations
Calculate
P
(
X
= 2)
Binomial distribution
= P(Spades) = 13/52
Slide2Binary
r.v
.
Outcomes of “X = 2”
(0, 0, 0, 1, 1)(0, 0, 1, 0, 1)(0, 0, 1, 1, 0)(0, 1, 0, 0, 1)(0, 1, 0, 1, 0)(0, 1, 1, 0, 0)(1, 0, 0, 0, 1)(1, 0, 0, 1, 0)(1, 0, 1, 0, 0)(1, 1, 0, 0, 0)
Random Sample
n
= 5(w/ or w/o replacement)
Random Samplen = 5(w/ replacement)
Infinite Populationof decks of 52 cards (Assume each is fair)
Calculate
P
(
X
= 2)
Finite
Population
N
= 52
(Assume fair)
Hypergeometric Distribution
10 combinations
Random Variable
X =
# Spades in sample
(x = 0, 1, 2, 3, 4, 5)
Binomial distribution
=
P
(Spades) = 13/52
Slide3Outcomes of “X = 2”(0, 0, 0, 1, 1)(0, 0, 1, 0, 1)(0, 0, 1, 1, 0)(0, 1, 0, 0, 1)(0, 1, 0, 1, 0)(0, 1, 1, 0, 0)(1, 0, 0, 0, 1)(1, 0, 0, 1, 0)(1, 0, 1, 0, 0)(1, 1, 0, 0, 0)
Random Samplen = 5(w/ replacement)
Hypergeometric Distribution
Finite
Population
N
= 52
(Assume fair)
Calculate
P
(X = 2)
10 combinations
Random Variable
X =
# Spades in sample
(x = 0, 1, 2, 3, 4, 5)
Binomial distribution = P(Spades) = 13/52
Random Sample
n = 5(w/o replacement)
Binary
r.v
.
Slide4Random Samplen = 5(w/o replacement)
?
?
?
?
?
?
Outcomes of “X = 2”
(0, 0, 0, 1, 1)
(0, 0, 1, 0, 1)
(0, 0, 1, 1, 0)
(0, 1, 0, 0, 1)
(0, 1, 0, 1, 0)(0, 1, 1, 0, 0)(1, 0, 0, 0, 1)(1, 0, 0, 1, 0)(1, 0, 1, 0, 0)(1, 1, 0, 0, 0)
Hypergeometric Distribution
Finite
Population
N
= 52
(Assume fair)
Calculate
P
(X = 2)
= 0.27428
Random Variable
X =
# Spades in sample
(
x
= 0, 1, 2, 3, 4, 5)
Binomial distribution
=
P
(Spades) = 13/52
Binary
r.v
.
Slide5Sample SpaceAll combinations of 5 cards from 52
#
Hypergeometric Distribution
Finite
Population
N
= 52
(Assume fair)
Calculate
P
(
X = 2)
Random Samplen = 5(w/o replacement)
# combinations of 2
Spades
from 13
Spades
# combinations of
3
Non-Spades
from 39
Non-Spades
= 0.27428
R command:
dhyper
(2, 13, 39, 5)
Random Variable
X =
# Spades in sample
(x = 0, 1, 2, 3, 4, 5)
x, s, N – s, n
Slide6Hypergeometric Distribution
Finite
Populationof size N
s = # Successes
N – s = # Failures
Random Sample
(without replacement) of size n N/10
Discrete random variableX = # Successes in sample(x = 0, 1, 2, 3, …,, n)
Then for any x = 0, 1, 2,…, the pmf p(x) is given by the following…
See textbook for
and 2.
# combinations of x Successes out of s Successes
# combinations of n – x Failures out of N – s Failures
# combinations of
n
out of
N
Slide7Hypergeometric Distribution
POPULATIONN = 100
s = 4 defectives
Random Sample(w/o replacement), n = 12
Discrete
random variable
X
= #
defectives
in sample
(
x
= 0
, 1, 2, 3
, 4)
Slide8Hypergeometric Distribution
POPULATIONN = 100
s = 4 defectives
Random Sample(w/o replacement), n = 12
Discrete
random variable
X
= #
defectives
in sample
(
x
= 0
, 1, 2, 3
, 4)
dhyper
(0:4, 4, 96, 12)
0 0.594684
1 0.335822
2 0.064431
3 0.004937
4 0.000126
Slide9X = # Successes (x) in n trials(x = 0, 1, 2, …, n) Classical Discrete ModelPopulation SizeSampling: replacement?Bernoulli trials?2pmf p(x)P(X = x)BinomialX ~ Bin(n, )dbinom(x, n, )∞with or withoutyesN < ∞withyesPoisson1X ~ Pois()dpois(x, )∞with or withoutyesHypergeometricX ~ Hyp(x, s, N, n)dhyper(x, s, N – s, n)N < ∞,n N/10withoutno
1 for rare events ONLY, i.e., small, n large2 independent outcomes, with constant = P(Success) in population
Slide10Negative Binomial Distribution
Discrete
random variable
X = # trials for s Successes(x = s, s+1, s+2, s+3, …)
See textbook for
and 2.
Random Sample(w/ or w/o replacement)
Infinite Populationof “Successes” and “Failures”
Then for any x = s, s+1, s+2,…, the pmf p(x) is…
P(Success) = P(Failure) = 1 –
Sample size NOT specified!
dnbinom(x–s, s, )
(In R,
X
counts the #
Failures
before
s
Successes)
Slide11Random Sample
(w/ replacement)
=
P
(Spades) = 13/52
Negative Binomial Distribution
Random Variable
X =
# trials for
s
= 6 Spades
(
x
= 6, 7, 8, 9, 10,…)
Slide12Random Sample
(w/ replacement)
=
P
(Spades) = 13/52
Negative Binomial Distribution
Random Variable
X =
# trials for s = 6 Spades(x = 6, 7, 8, 9, 10,…)
SPECIAL CASE OF NEGATIVE BINOMIAL: s
= 1
Random Variable
X = # trials for s = 1 Spades(x = 1, 2, 3, 4, 5,…)
Slide13Random VariableX = # trials for s = 1 Spades(x = 1, 2, 3, 4, 5,…)
Random Sample
(w/ replacement)
=
P
(Spades) = 13/52
Negative Binomial Distribution
SPECIAL CASE OF NEGATIVE BINOMIAL: s
= 1
(
x
– 1) Failures
1 Success!
Geometric Distribution
dgeom
(x–1, )
(In R,
X
counts the #
Failures
before
1
Success)
Slide14Random VariableX = # trials for s = 1 Spades(x = 1, 2, 3, 4, 5,…)
Random Sample
(w/ replacement)
=
P
(Spades) = 13/52
Negative Binomial Distribution
SPECIAL CASE OF NEGATIVE BINOMIAL: s
= 1
Geometric Distribution
Slide15Random VariableX = # trials for s = 1 Spades(x = 1, 2, 3, 4, 5,…)
Random Sample(w/ replacement)
= P(Spades) = 13/52
Negative Binomial Distribution
Geometric Distribution
x
1
2
34567p(x).250.188.141.105.079.059.044
ExerciseGraph cdf
F(x)
Slide16Random VariableX = # trials for s = 1 Spades(x = 1, 2, 3, 4, 5,…)
Random Variables X1 = # Spades X2 = # Clubs X3 = # Hearts X4 = # DiamondsFor i = 1, 2, 3, 4 xi = 0, 1, 2,…,10 with x1 + x2 + x3 + x4 = 10.
Geometric Distribution
Random Sample(w/ replacement)
=
P
(Spades) = 13/52
Multinomial Distribution
1
=
P
(Spades) = 13/52
2 = P(Clubs) = 13/523 = P(Hearts) = 13/524 = P(Diamonds) = 13/52
Random Samplen = 10(w/ replacement)
Slide17Discrete random variableX = # “Successes” in sample(n – X = # “Failures” in sample)(0, 1, 2, 3, …, n)
RANDOM SAMPLE of n “Bernoulli trials”
BINARY POPULATIONof “Successes” vs. “Failures” P(Success) = P(Failure) = 1 –
Then
X is said to follow a Binomial distribution, written X ~ Bin(n, ), with “probability mass function” p(x) = … x = 0, 1, 2, …, n
RECALL…
Slide18Discrete random variableX = # “Successes” in sample(n – X = # “Failures” in sample)(0, 1, 2, 3, …, n)
BINARY POPULATIONof “Successes” vs. “Failures” P(Success) = P(Failure) = 1 –
Then X is said to follow a Binomial distribution, written X ~ Bin(n, ), with “probability mass function” p(x) = … x = 0, 1, 2, …, n
RECALL…
RANDOM SAMPLE
of
n “Bernoulli trials”
OR…
Slide19Discrete random variableX = # “Successes” in sample(n – X = # “Failures” in sample)(0, 1, 2, 3, …, n)
Discrete random variablesX1 = # “Successes” in sampleX2 = # “Failures” in sample = n – X1(0, 1, 2, 3, …, n)
BINARY POPULATIONof “Successes” vs. “Failures” P(Success) = 1 P(Failure) = 2
1 + 2 = 1
RECALL…
OR…
Then
X is said to follow a Binomial distribution, written X ~ Bin(n, ), with “probability mass function” p(x) = … x = 0, 1, 2, …, n
RANDOM SAMPLE
of
n
“Bernoulli trials”
Slide20BINARY POPULATION
of “Successes” vs. “Failures” P(Success) = 1 P(Failure) = 2
1 + 2 = 1
RECALL…
RANDOM SAMPLE of n “Bernoulli trials”
OR…
Discrete random variablesX1 = # “Successes” in sampleX2 = # “Failures” in sample = n – X1(0, 1, 2, 3, …, n)
Then
X1 and X2 “jointly” follow a Binomial distribution, written X1 ~ Bin(n, 1), X2 ~ Bin(n, 2), with “probability mass function”… for any x1 = 0, 1, 2, …, n, x2 = 0, 1, 2,…, n, with .
x1 + x2 = n
Slide21Then the components of
X = (X1, X2,…, Xk) “jointly” follow a Multinomial distribution, written X ~ Multi(n, 1, 2, …, k), with “probability mass function” p(x1,…, xk) = for any xi = 0, 1, 2, …, n, with .
1
+ 2 = 1
BINARY POPULATIONof “Successes” vs. “Failures” P(Success) = 1 P(Failure) = 2
1 + 2 + … + k = 1
POPULATION
of k categories P(Category 1) = 1 P(Category 2) = 2 P(Category k) = k
Discrete
random variablesX1 = # “Successes” in sampleX2 = # “Failures” in sample = n – X1(0, 1, 2, 3, …, n)
Discrete
random variablesX1 = # Category 1 in sampleX2 = # Category 2 in sampleXk = # Category k in sample
RANDOM SAMPLE
of
n
“Bernoulli trials”
Slide22Multinomial Distribution
Random Sample(w/ replacement)
=
P
(Spades) = 13/52
Random Variable
X =
# trials for
s = 6 Spades(x = 6, 7, 8, 9, 10,…)
Random Variables
X
1
= # Spades X2 = # Clubs X3 = # Hearts X4 = # Diamondsxi = 0, 1, 2,…,10 with x1 + x2 + x3 + x4 = 10.
1 = P(Spades) = 13/522 = P(Clubs) = 13/523 = P(Hearts) = 13/524 = P(Diamonds) = 13/52
Random Samplen = 10(w/ replacement)
dmultinom
(c(1,2,3,4), 10,
c(.25,.25,.25,.25))