Probability

Probability

Multiplication Rule of Probability

  • For indenpendent events, $P(A \cap B) = P(A)P(B)$.
  • For dependent events, $P(A \cap B) = P(B)P(A B)$.

Conditional Probability

1. Bayes’ rule

  • The conditional probability of event A given event B is given by
\[P(A|B)=\frac{P(B|A)P(A)}{P(B)}\]
  • If $P(A B)=P(A)$, then A and B are independent, since knowing about B tells nothing about the probability of A having also occurred.
  • If $P(A \cap B C) = P(A C)P(B C)$, A and B are conditionally independent given the occurrence of C.

2. Law of total probability

  • Given several disjoint events with B having occurred, the probability of an event A having also occurred can be broken down with \(P(A) = P(A|B_1)P(B_1)+...+P(A|B_n)P(B_n)\)

Probability Distributions

Binomial distribution

  • The binomial distribution is the discrete probability distribution of the number of successes in a sequence of $n$ independent experiments, each asking a yes-no question, and each with its own boolean value outcome: success (with probability $ = p$) or failure (with probability $ = 1 - p$).
  • For a single trial / experiment, the binomial distribution is a Bernoulli distribution.
  • The binomial distribution gives the probability of $k$ number of successes in $n$ independent trials, where each trial has probability of $p$ of success. Its probability mass function (PMF) is given by

\(P(X = k) = \binom{n}{k}p^k(1 - p)^{n - k}\)

  • The mean and variance are: $\mu = np$, $\sigma^2 = np(1-p)$.

References