Proof: Law of total probability and Bayes’ theorem

Author

Sophie Chowgule

Summary
This proof sheet demonstrates that the law of total probability and Bayes’ theorem are true.

Before reading this proof sheet, it is recommended that you read Guide: Conditional probability and Guide: Law of total probability and Bayes’ theorem.

Proof of the law of total probability

First of all, here is a restatement of the law of total probability from Guide: Law of total probability and Bayes’ theorem:

Definition of the law of total probability

Suppose an event \(B\) depends on several possible scenarios. These scenarios can be described by events \(A_1,A_2,\ldots, A_n\), that are:

  • Mutually exclusive: they cannot occur at the same time, and
  • Exhaustive: one of them must always occur.

Then, the law of total probability states that the probability of event \(B\) is:

\[ \mathbb{P}(B) = \sum_{i=1}^{n} \mathbb{P}(A_i) \mathbb{P}(B \mid A_i) \]

The proof of the law of total probability comes directly from the definition of conditional probability given in Guide: Conditional probability:

\[ \mathbb{P}(B \mid A_i) = \frac{\mathbb{P}(B \cap A_i)}{\mathbb{P}(A_i)} \]

Multiplying by \(\mathbb{P}(A_i)\) gives the multiplication rule (again from Guide: Conditional probability) :

\[ \mathbb{P}(B \cap A_i) = \mathbb{P}(B \mid A_i) \mathbb{P}(A_i) \]

As scenarios \(A_1, A_2, \dots, A_n\) are mutually exclusive (so \(A_i \cap A_j = \emptyset\) for all \(1\leq i\neq j \leq n\)) and exhaustive (\(\bigcup_{1\leq i \leq n}A_i = B\)), it follows from results in set theory (see [Guide: Operations on sets]) that:

\[ \mathbb{P}(B) = \mathbb{P}(B \cap A_1) + \mathbb{P}(B \cap A_2) + \dots + \mathbb{P}(B \cap A_n) = \sum_{i=1}^n \mathbb{P}(B \cap A_i) \]

Substituting the above expressions gives:

\[ \mathbb{P}(B) = \mathbb{P}(B \mid A_1) \mathbb{P}(A_1) + \mathbb{P}(B \mid A_2) \mathbb{P}(A_2) + \dots + \mathbb{P}(B \mid A_n) \mathbb{P}(A_n) = \sum_{i=1}^n \mathbb{P}(B\mid A_i)\mathbb{P}(A_i) \]

Which results in the law of total probability:

\[ \mathbb{P}(B) = \sum_{i=1}^{n} \mathbb{P}(A_i) \mathbb{P}(B \mid A_i) \]

Proof of Bayes’ theorem

Here is the statement of Bayes’ theorem from Guide: Law of total probability and Bayes’ theorem:

Statement of Bayes’ Theorem

If \(A\) and \(B\) are events with \(\mathbb{P}(B) > 0\), then Bayes’ Theorem states:

\[ \mathbb{P}(A \mid B) = \dfrac{\mathbb{P}(B \mid A) \mathbb{P}(A)}{\mathbb{P}(B)}. \]

where:

  • \(\mathbb{P}(A \mid B)\) is the probability of \(A\) given \(B\),
  • \(\mathbb{P}(B \mid A)\) is the probability of \(B\) given \(A\),
  • \(\mathbb{P}(A)\) and \(\mathbb{P}(B)\) are the individual probabilities of \(A\) and \(B\), respectively.

Bayes’ Theorem is derived directly from the definition of conditional probability: see Guide: Conditional probability. Start with the conditional probabilities of two events \(A\) and \(B\):

\[ (1)\quad \mathbb{P}(A \mid B) = \frac{\mathbb{P}(A \cap B)}{\mathbb{P}(B)} \quad\textsf{ and }\quad (2)\quad \mathbb{P}(B \mid A) = \frac{\mathbb{P}(A \cap B)}{\mathbb{P}(A)} \]

You can rearrange \((2)\) by multiplying both sides by \(\mathbb{P}(A)\), giving the multiplication rule:

\[ \mathbb{P}(A \cap B) = \mathbb{P}(B \mid A) \mathbb{P}(A) \]

Substitute this result into equation \((1)\) to get:

\[ \mathbb{P}(A \mid B) = \frac{\mathbb{P}(B \mid A) \mathbb{P}(A)}{\mathbb{P}(B)} \]

This gives Bayes’ Theorem, a way to reverse conditional probabilities when direct calculation is difficult.

Further reading

Click this link to go back to Guide: Law of total probability and Bayes’ theorem.

Version history

v1.0: initial version created 04/25 by Sophie Chowgule as part of a University of St Andrews VIP project.

This work is licensed under CC BY-NC-SA 4.0.

Mailing List



Feedback

Your feedback is appreciated and useful. Feel free to leave a comment here,
but please be specific with any issues you encounter so we can help to resolve them
(for example, what page it occured on, what you tried, and so on).