Prob

Probability Fact Sheets
Combinatorics
The number of ways of choosing k objects from n, with repetitions, where

the order counts, is nk .
The number of ways of choosing k objects from n, without repetitions, where
the order counts is
n!
= n(n − 1) · · · n − (k − 1).
(n − k)!
This is called the number of permutations of k objects from n.
The number of ways of choosing k objects from n, without repetititions,
where the order doesn’t count, is

n n! n(n − 1) · · · n − (k − 1)
= =
k k!(n − k)! k(k − 1) · · · 1
This is called the number of combinations of k objects from n, and happens
to be the kth entry of the nth row of Pascal’s triangle (assuming you start
counting with 0).
The binomial theorem:
Xn
n n i n−i
(x + y) = xy
i
i=0

n n n−1 n n−2 2 n
= x + x y+ x y + ... xy n−1 + y n
1 2 n−1
Probability Measures
Let X be the set of possible outcomes of an experiment. Subsets of X are

called events. If A and B are events, then
• A ∪ B is the union of the two events (that is, the outcomes that are in
either A or B)
• A ∩ B is the intersection of the two events (that is, the outcomes that
are in both A and B)
• Ā is the complement of A (that is, the outcomes that are not in A)
Events A and B are incompatible (or “mutually exclusive,” or “disjoint”) if
A ∩ B = ∅, that is, there is no outcome common to both events.
Definition: A (finitely additive) probability measure on X is a function P
that assigns a real number between 0 and 1 to subsets of X (that is, elements
of some field of subsets of X), satisfying the following:
• P (X) = 1
• If A and B are incompatible events, then P (A ∪ B) = P (A) + P (B)
These facts follow from the definition:

• P (∅) = 0
• If A ⊆ B then P (A) ≤ P (B)
Suppose X is finite, and each of the outcomes in X is equally likely. Then

for any event A,
|A|
P (A) = ,
|X|
where |A| is the number of elements in A, and |X| is the number of elements
in X.
Bernoulli’s Theorem
Two events A and B are said to be independent if
P (A ∩ B) = P (A)P (B).
Suppose one runs n independent trials of an experiment in which the prob-

ability of success in each trial is p and the probability of failure is q = 1 − p.
Then the probability that there are exactly k successes in these n trials is

n k n−k
p q .
k
Bernoulli’s Theorem: Suppose one repeatedly runs independent trials of an

experiment in which the probability of success in each trial is p. Let A and
B be small positive numbers. Then there is a value of n large enough so
that the probability that the ratio of successes in n trials is not within A of
p is less than B.
In other words: if you run the experiment long enough, the fraction of
successes is very likely to be close to the “correct” probability. See the
discussion on page 203 of TTT.
Bayes’ Theorem
Let X be a (finite) probability space, and let A and B be events. The

probability of A given B is
P (A ∩ B)
P (A|B) =
P (B)
Multiplying through, we see that
P (A ∩ B) = P (A|B)P (B) = P (B|A)P (A).
Let A and B be any two events. Since A = (A ∩ B) ∪ (A ∩ B̄), and (A ∩ B)

and (A ∩ B̄) are incompatible, we have
P (A) = P (A ∩ B) + P (A ∩ B̄)
= P (A|B)P (B) + P (A|B̄)P (B̄).
Now let H and E be events. Putting all the information together, we have
Bayes’ theorem:
P (H ∩ E)
P (H|E) =
P (E)
P (E|H)P (H)
=
P (E|H)P (H) + P (E|H̄)P (H̄)
Interpret this as follows:

• Let H be some scientific hypothesis you want to test.
• Let P (H) be your a priori estimate of the probability that the hy-
pothesis is correct (before running an experiment).
• Then P (H̄) your estimate of the probability that the hypothesis is

false.
• After running an experiment, you find some evidence, E.
• Then P (E|H) is the probability that the experiment would turn out
this way, assuming your hypothesis is correct.
• And P (E|H̄) is the probability that the experiment would turn out
this way, if your hypothesis is incorrect.
Bayes’ theorem then allows you to calculate the probability that the hy-
pothesis is correct, given the evidence you have gathered.
In the discussion on pages 206-210 of TTT, the “hypothesis” is that the
probability of success in each trial of the experiment is somewhere between
a and b, and the “evidence” is the fraction of successes in some large number
of trials. The analysis shows that if the experiment returns a large num-
ber of sucesses and no failures, then the only likely hypothesis is that the
probability of success is very close to 1.

Prob

Uploaded by

Copyright:

Available Formats

Prob

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Prob

Uploaded by

Copyright:

Available Formats

Probability Fact Sheets

The number of ways of choosing k objects from n, with repetitions, where

Let X be the set of possible outcomes of an experiment. Subsets of X are

• If A and B are incompatible events, then P (A ∪ B) = P (A) + P (B)

These facts follow from the definition:

• If A ⊆ B then P (A) ≤ P (B)

Suppose X is finite, and each of the outcomes in X is equally likely. Then

Two events A and B are said to be independent if

Suppose one runs n independent trials of an experiment in which the prob-

Bernoulli’s Theorem: Suppose one repeatedly runs independent trials of an

Let X be a (finite) probability space, and let A and B be events. The

P (A ∩ B) = P (A|B)P (B) = P (B|A)P (A).

Let A and B be any two events. Since A = (A ∩ B) ∪ (A ∩ B̄), and (A ∩ B)

Interpret this as follows:

• Then P (H̄) your estimate of the probability that the hypothesis is

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.