![Page 1: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/1.jpg)
Probability Theory:Counting in Terms of Proportions
Great Theoretical Ideas In Computer Science
Steven Rudich, Anupam Gupta CS 15-251 Spring 2005
Lecture 20 March 24, 2005 Carnegie Mellon University
![Page 2: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/2.jpg)
A Probability Distribution
HEIGHT
Proportion of
MALES
50% of the male
population
75% of the male
population
50% of the male
population
![Page 3: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/3.jpg)
The Descendants Of Adam
Adam was X inches tall.
He had two sonsOne was X+1 inches tallOne was X-1 inches tall
Each of his sons had two sons …
![Page 4: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/4.jpg)
X
X-1 X+1
1 1
X-2 X+2X
X-3 X+3X-1 X+1
X-4 X+4X-2 X+2X
156 6
5 510 10
1520
![Page 5: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/5.jpg)
1
X-1 X+1
1 1
X-2 X+2X
X-3 X+3X-1 X+1
X-4 X+4X-2 X+2X
156 6
5 510 10
1520
![Page 6: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/6.jpg)
1
1 1
1 1
X-2 X+2X
X-3 X+3X-1 X+1
X-4 X+4X-2 X+2X
156 6
5 510 10
1520
![Page 7: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/7.jpg)
1
1 1
1 1
1 12
X-3 X+3X-1 X+1
X-4 X+4X-2 X+2X
156 6
5 510 10
1520
![Page 8: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/8.jpg)
1
1 1
1 1
1 12
1 13 3
X-4 X+4X-2 X+2X
156 6
5 510 10
1520
![Page 9: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/9.jpg)
1
1 1
1 1
1 12
1 13 3
1 14 46
156 6
5 510 10
1520
![Page 10: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/10.jpg)
1
1 1
1 1
1 12
1 13 3
1 14 46
156 6
5 510 10
1520
In nth generation, there will be 2n males, each with one of n+1 different heights:
h0< h1 < . . .< hn.
hi = (X – n + 2i) occurs with proportion
![Page 11: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/11.jpg)
Unbiased Binomial Distribution On n+1 Elements.
Let S be any set {h0, h1, …, hn} where each element hi has an associated probability
Any such distribution is called an Unbiased Binomial Distribution or an
Unbiased Bernoulli Distribution.
![Page 12: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/12.jpg)
As the number of elements gets larger, the shape of the unbiased
binomial distribution converges to a Normal (or Gaussian) distribution.
Mean
![Page 13: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/13.jpg)
As the number of elements gets larger, the shape of the unbiased
binomial distribution converges to a Normal (or Gaussian) distribution.
¡2m
m¡ t
¢
¡2mm
¢ ≈ e¡ (t2=m)
![Page 14: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/14.jpg)
1 14 46
Coin Flipping in Manhattan
At each step, we flip a coin to decide which way to go.
What is the probability of ending at the intersection of
Avenue i and Street (n-i) after n steps?
![Page 15: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/15.jpg)
1 14 46
Coin Flipping in Manhattan
At each step, we flip a coin to decide which way to go.
What is the probability of ending at the intersection of
Avenue i and Street (n-i) after n steps?
![Page 16: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/16.jpg)
Coin Flipping in Manhattan
At each step, we flip a coin to decide which way to go.
What is the probability of ending at the intersection of
Avenue i and Street (n-i) after n steps?
1 14 46
![Page 17: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/17.jpg)
Coin Flipping in Manhattan
At each step, we flip a coin to decide which way to go.
What is the probability of ending at the intersection of
Avenue i and Street (n-i) after n steps?
1 14 46
![Page 18: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/18.jpg)
Coin Flipping in Manhattan
At each step, we flip a coin to decide which way to go.
What is the probability of ending at the intersection of
Avenue i and Street (n-i) after n steps?
1 14 46
![Page 19: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/19.jpg)
Coin Flipping in Manhattan
2n different paths to level n, each equally likely.
The probability of i heads occurring on the path we generate is:
1 14 46
![Page 20: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/20.jpg)
n-step Random Walk on a line
Start at the origin: at each point, flip an unbiased coin to decide whether to go right or left.
The probability that, in n steps, we take i steps to the right and n-i to the left (so we are at position 2i-n) is:
1 14 46
![Page 21: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/21.jpg)
n-step Random Walk on a line
Start at the origin: at each point, flip an unbiased coin to decide whether to go right or left.
The probability that, in n steps, we take i steps to the right and n-i to the left (so we are at position 2i-n) is:
1 14 46
![Page 22: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/22.jpg)
n-step Random Walk on a line
Start at the origin: at each point, flip an unbiased coin to decide whether to go right or left.
The probability that, in n steps, we take i steps to the right and n-i to the left (so we are at position 2i-n) is:
1 14 46
![Page 23: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/23.jpg)
n-step Random Walk on a line
Start at the origin: at each point, flip an unbiased coin to decide whether to go right or left.
The probability that, in n steps, we take i steps to the right and n-i to the left (so we are at position 2i-n) is:
1 14 46
![Page 24: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/24.jpg)
n-step Random Walk on a line
Start at the origin: at each point, flip an unbiased coin to decide whether to go right or left.
The probability that, in n steps, we take i steps to the right and n-i to the left (so we are at position 2i-n) is:
1 14 46
![Page 25: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/25.jpg)
Again, a Normal or Gaussian distribution!
Let’s look after n steps
23
pn
50% oftime
![Page 26: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/26.jpg)
Again, a Normal or Gaussian distribution!
Let’s look after n steps
68% oftime
pn
![Page 27: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/27.jpg)
Again, a Normal or Gaussian distribution!
Let’s look after n steps
95% oftime
2p
n
![Page 28: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/28.jpg)
Again, a Normal or Gaussian distribution!
Let’s look after n steps
99.7% oftime
3p
n
![Page 29: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/29.jpg)
Probabilities and Countingare intimately related ideas…
![Page 30: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/30.jpg)
Probabilities and counting
Say we want to count the number of X's with property P
One way to do it is to ask "if we pick an X at random,
what is the probability it has property P?" and then multiply by the number of X's.
(# of X with prop. P)
(total # of X)
Probability of X with prop. P
=
![Page 31: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/31.jpg)
Probabilities and counting
Say we want to count the number of X's with property P
One way to do it is to ask "if we pick an X at random,
what is the probability it has property P?" and then multiply by the number of X's.
(# of X with prop. P)
(total # of X)
Probability of X with prop. P
=×
![Page 32: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/32.jpg)
How many n-bit strings have an even number of 1’s?
If you flip a coin n times, what is the probability you get an even number of heads? Then multiply by 2n.
Say prob was q after n-1 flips.
Then, after n flips it is ½q + ½(1-q) = ½.
(# of X with prop. P)(total # of X)
Probability of X with prop. P
=×
![Page 33: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/33.jpg)
1 14 46
Binomial distribution with bias p
Start at the top. At each step, flip a coin with a bias p of heads to decide which way to go.
What is the probability of ending at the intersection of
Avenue i and Street (n-i) after n steps?
p1-p
![Page 34: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/34.jpg)
1 14 46
Binomial distribution with bias p p1-p
Start at the top. At each step, flip a coin with a bias p of heads to decide which way to go.
What is the probability of ending at the intersection of
Avenue i and Street (n-i) after n steps?
![Page 35: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/35.jpg)
1 14 46
Binomial distribution with bias p p
p1-p
1-pp
Start at the top. At each step, flip a coin with a bias p of heads to decide which way to go.
What is the probability of ending at the intersection of
Avenue i and Street (n-i) after n steps?
![Page 36: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/36.jpg)
1 14 46
Binomial distribution with bias p p
p1-p
1-pp
Start at the top. At each step, flip a coin with a bias p of heads to decide which way to go.
The probability of any fixed path with i heads (n-i tails) being chosen is: pi (1-p)n-i
Overall probability we get i heads is:
![Page 37: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/37.jpg)
Bias p coin flipped
n times. Probability
of exactly i heads
is:
![Page 38: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/38.jpg)
How many n-trit strings have even number of 0’s?
If you flip a bias 1/3 coin n times, what is the probability qn you get an even number of heads? Then multiply by 3n. [Why is this right?]
Then q0=1.
Say probability was qn-1 after n-1 flips.
Then, qn = (2/3)qn-1 + (1/3)(1-qn-1).
Rewrite as: qn – ½ = 1/3(qn-1- ½)
pn = qn – ½
pn = 1/3 pn-1
and p0 = ½.
So, qn – ½ = (1/3)n ½. Final count = ½ + ½3n
![Page 39: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/39.jpg)
Some puzzles
![Page 40: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/40.jpg)
Teams A and B are equally good.
In any one game, each is equally likely to win.
What is most likely length of a “best of 7” series?
Flip coins until either 4 heads or 4 tails. Is this more likely to take 6 or 7 flips?
![Page 41: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/41.jpg)
Actually, 6 and 7 are equally likely
To reach either one, after 5 games, it must be 3 to 2.
½ chance it ends 4 to 2. ½ chance it doesn’t.
![Page 42: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/42.jpg)
Another view
1 1
156 6
5 510 10
1520
WW
L
LW
4
5
6
7 7
6
5
4L
3 to 24 to 2
3-3 tie 7 games
![Page 43: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/43.jpg)
One bag has two silver coins, another has two gold coins, and the third has one of each.
One of the three bags is selected at random. Then one coin is selected at random from the two in the bag. It turns out to be gold.
What is the probability that the other coin is gold?
Silver and Gold
![Page 44: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/44.jpg)
3 choices of bag2 ways to order bag contents
6 equally likely paths.
X
![Page 45: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/45.jpg)
X
X X X
Given you see a , 2/3 of remaining paths have a second gold.
![Page 46: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/46.jpg)
So, sometimes, probabilities can be counter-intuitive
![Page 47: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/47.jpg)
Language Of Probability
The formal language of probability is a
very important tool in describing and
analyzing probability distributions.
![Page 48: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/48.jpg)
Finite Probability Distribution
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
The weights must satisfy:
![Page 49: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/49.jpg)
Finite Probability Distribution
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
For notational convenience we will define D(x) = p(x).
S is often called the sample spaceand elements x in S are called samples.
![Page 50: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/50.jpg)
Sample space
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
SSample space
![Page 51: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/51.jpg)
Probability
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
S
x
D(x) = p(x) = 0.2weight or probability
of x
![Page 52: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/52.jpg)
Probability Distribution
S
0.2
0.13
0.06
0.110.17
0.10.13
0
0.1
weights must sum to 1
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
![Page 53: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/53.jpg)
Events
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
Any set E ½ S is called an event. The probability of event E is
![Page 54: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/54.jpg)
S
Event E
Events
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
![Page 55: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/55.jpg)
S0.17
0.10.13
0
PrD[E] = 0.4
Events
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
![Page 56: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/56.jpg)
Uniform Distribution
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
If each element has equal probability, the distribution is said to be uniform.
![Page 57: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/57.jpg)
S1/9
Each p(x) = 1/9.
1/9
1/9
1/9 1/9
1/9
1/9
1/9
1/9
Uniform Distribution
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
![Page 58: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/58.jpg)
PrD[E] = |E|/|S| = 4/9
Uniform Distribution
S
A (finite) probability distribution D is a finite set S of elements, where each element x in S has a positive real weight, proportion, or probability p(x).
![Page 59: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/59.jpg)
A fair coin is tossed 100 times in a row.
What is the probability that we get exactly half heads?
![Page 60: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/60.jpg)
The sample space S is the set of all outcomes {H,T}100.
Using the Language
A fair coin is tossed 100
times in a row.
Each sequence in S is equally likely, and hence has probability 1/|S|=1/2100.
![Page 61: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/61.jpg)
Using the Language: visually
A fair coin is tossed 100
times in a row.
S = all sequencesof 100 tosses
x
x = HHTTT……THp(x) = 1/|S|
Uniform distribution!
![Page 62: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/62.jpg)
A fair coin is tossed 100 times in a row.
What is the probability that we get exactly half heads?
OK
![Page 63: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/63.jpg)
The event that we see half heads is
E = {x 2 S | x has 50 heads}
Using the Language
Probability of exactly half
tails?
100
50E
And
![Page 64: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/64.jpg)
Picture
Set of all 2100 sequences{H,T}100
Probability of event E = proportion of E in S 100
100
50
2
E
S
Event E = Set of sequences with 50 H’s and 50 T’s
![Page 65: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/65.jpg)
Using the Language
Probability of exactly half
tails?
Answer:
Pr[E] = |E|/|S|=|E|/2100
![Page 66: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/66.jpg)
Suppose we roll a white die and a black die.
What is the probability that sum is 7 or 11?
![Page 67: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/67.jpg)
Event E = all (x,y) pairs with x+y = 7 or 11
Same methodology!
Sample space S = { (1,1), (1,2), (1,3), (1,4), (1,5), (1,6), (2,1), (2,2), (2,3), (2,4), (2,5), (2,6), (3,1), (3,2), (3,3), (3,4), (3,5), (3,6), (4,1), (4,2), (4,3), (4,4), (4,5), (4,6), (5,1), (5,2), (5,3), (5,4), (5,5), (5,6), (6,1), (6,2), (6,3), (6,4), (6,5), (6,6) }
Pr[E] = |E|/|S| = proportion of E in S = 8/36
Pr(x) = 1/368 x 2 S
![Page 68: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/68.jpg)
23 people are in a room.
Suppose that all possible assignments of birthdays to the
23 people are equally likely.
What is the probability that two people will have the same
birthday?
![Page 69: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/69.jpg)
Event E = {x 2 | two numbers in x are same}What is |E| ?
And the same methods again!
Sample space = { 1, 2, 3, …, 366}23
Pretend it’s always a leap year
x = (17,42,363,1,…, 224,177)
Ecount instead!
23 numbers
![Page 70: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/70.jpg)
all sequences in that have no repeated numbers
E
.51E
![Page 71: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/71.jpg)
Another way to calculate Pr(no collision)
Pr(1st person doesn’t collide) = 1.Pr(2nd doesn’t | no collisions yet) = 365/366.Pr(3rd doesn’t | no collisions yet) = 364/366.Pr(4th doesn’t | no collisions yet) = 363/366.…Pr(23rd doesn’t| no collisions yet) = 344/366.
1
365/366
364/366
363/366
![Page 72: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/72.jpg)
The probability of event A given event B is written Pr[ A | B ]
and is defined to be =
Pr
Pr
A B
B
More Language Of Probability
A
Bproportion of A B
to B
![Page 73: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/73.jpg)
Suppose we roll a white die and black die.
What is the probability that the white is 1
given that the total is 7?
event A = {white die = 1}
event B = {total = 7}
![Page 74: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/74.jpg)
event A = {white die = 1} event B = {total = 7}
Sample space S = { (1,1), (1,2), (1,3), (1,4), (1,5), (1,6), (2,1), (2,2), (2,3), (2,4), (2,5), (2,6), (3,1), (3,2), (3,3), (3,4), (3,5), (3,6), (4,1), (4,2), (4,3), (4,4), (4,5), (4,6), (5,1), (5,2), (5,3), (5,4), (5,5), (5,6), (6,1), (6,2), (6,3), (6,4), (6,5), (6,6) }
|A \ B| = Pr[A | B] = Pr[A \ B] = 1/36
|B| Pr[B] 1/6
This way does not care about the distribution.
Can do this because is uniformly distributed.
Pr[A | B]
![Page 75: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/75.jpg)
Independence!
A and B are independent events if
A
B
Pr[ A | B ] = Pr[ A ]
Pr[ A \ B ] = Pr[ A ] Pr[ B ]
Pr[ B | A ] = Pr[ B ]
What about Pr[ A | not(B) ] ?
![Page 76: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/76.jpg)
Independence!
A1, A2, …, Ak are independent events if knowingif some of them occurred does not change
the probability of any of the others occurring.
Pr[A1 | A2 A3] = Pr[A1]
Pr[A2 | A1 A3] = Pr[A2]
Pr[A3 | A1 A2] = Pr[A3]
Pr[A1 | A2 ] = Pr[A1] Pr[A1 | A3 ] = Pr[A1]
Pr[A2 | A1 ] = Pr[A2] Pr[A2 | A3] = Pr[A2]
Pr[A3 | A1 ] = Pr[A3] Pr[A3 | A2] = Pr[A3]
E.g., {A_1, A_2, A_3}are independent
events if:
![Page 77: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/77.jpg)
Independence!
A1, A2, …, Ak are independent events if knowingif some of them occurred does not change
the probability of any of the others occurring.
Pr[A|X] = Pr[A]
8 A 2 {Ai} 8 X a conjunction of any of the others (e.g., A2 and A6 and A7)
![Page 78: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/78.jpg)
One bag has two silver coins, another has two gold coins, and the third has one of each.
One of the three bags is selected at random. Then one coin is selected at random from the two in the bag. It turns out to be gold.
What is the probability that the other coin is gold?
Silver and Gold
![Page 79: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/79.jpg)
Let G1 be the event that the first coin is gold.
Pr[G1] = 1/2
Let G2 be the event that the second coin is gold.
Pr[G2 | G1 ] = Pr[G1 and G2] / Pr[G1]
= (1/3) / (1/2)
= 2/3
Note: G1 and G2 are not independent.
![Page 80: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/80.jpg)
Monty Hall problem
•Announcer hides prize behind one of 3 doors at random.
•You select some door.
•Announcer opens one of others with no prize.
•You can decide to keep or switch.
What to do?
![Page 81: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/81.jpg)
Monty Hall problem
•Sample space = { prize behind door 1, prize behind door 2, prize behind door 3 }.
Each has probability 1/3.
Stayingwe win if we choose
the correct door
Switchingwe win if we choose
the incorrect door
Pr[ choosing correct door ] = 1/3.
Pr[ choosing incorrect door ] = 2/3.
![Page 82: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/82.jpg)
why was this tricky?
We are inclined to think:
“After one door is opened, others are equally likely…”
But his action is not independent of yours!
![Page 83: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/83.jpg)
Random walks and electrical networks
-
What is chance I reach yellow before magenta?
Same as voltage if edges are resistors and we put
1-volt battery between yellow and magenta.
![Page 84: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/84.jpg)
Random walks and electrical networks
-
• px = Pr(reach yellow first starting from x)
• pyellow= 1, pmagenta = 0, and for the rest,
• px = Averagey2 Nbr(x)(py)
Same as equations for voltage if edges all have same resistance!
![Page 85: Probability Theory: Counting in Terms of Proportions](https://reader035.vdocuments.us/reader035/viewer/2022070404/56813c1d550346895da59358/html5/thumbnails/85.jpg)
Random walks come up all the time
•Model stock as: each day has 50/50 chance of going up by $1, or down by $1.
•If currently $k, what is chance will reach $100 before $0?
•Ans: k/100.
•Will see other ways of analyzing later…
1 1
6
5 510 10