Scribd
Upload
12 views2 pages

Bayesian Rule: Defective Item Example

The document discusses how Bayesian rule can be extended to cases with more than two mutually exclusive and exhaustive events. It provides a formula to calculate the revised probability of an event occurring after observing another event. As an example, it calculates the probability that an item came from a specific factory given it was found to be defective, using the percentages produced and defective from each factory and Bayes' theorem.

Uploaded by

aman_nsu
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
12 views2 pages

Bayesian Rule: Defective Item Example

The document discusses how Bayesian rule can be extended to cases with more than two mutually exclusive and exhaustive events. It provides a formula to calculate the revised probability of an event occurring after observing another event. As an example, it calculates the probability that an item came from a specific factory given it was found to be defective, using the percentages produced and defective from each factory and Bayes' theorem.

Uploaded by

aman_nsu
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Bayesian rule can be extended to the case of more than two types of events.

Suppose we have mutually exclusive and exhaustive events A, B, C. Also suppose


D is such that event D can occur with A or B or C. Suppose we know P(D/A), P
(D/B) and P(D/C). After an experiment it is found that D occurred. Can we use this
information to revise the probabilities of A, B and C? Yes by using the extended
version of the Bayesian rule. I will show the formula below:
P(A/D) can be found by
( ) (

) / ( ) (

( ) (

( ) (

Similarly we can find the revised probabilities of B or C.


Example 2 of Bayesian Rule: An item is produced in 3 different factories, C1, C2,
C3. The proportions produced in the 3 factories and the proportions defective in
each are as follows:
Factory

% produced

% defective

C1

50

C2

30

4
An item is purchased and found to be defective. Let us call this event D. After this
result, what is the probability that it was from factory C1?
C3

20

Answer: The probability asked by the question (the posterior probability after the
result of defective is known) can be indicated in symbols as P(C1/D). Note that the
last column shows the conditional probability of defective for each factory. For
example, P(D/C1) is 0.02 or 2 percent. Thus we see that the question asks a reverse,
conditionality to what is given. Clearly this is a case of Bayes Theorem.
The probability of a defective P(D) before the result is known can be found from
the Law of Total probability or the rule of addition for mutually exclusive
events. The defective item can come from any of the factories. Therefore, its total
probability is the sum of the joint probabilities with the three factories.
The joint probability of D with C1 is P(D and C1) = P(C1)*P(D/C1) = 0.5*0.02 =
0.01;

The joint probability of D with C2 is P(D and C2) = P(C2)*P(D/C2) = 0.3*0.03 =


0.009; and
The joint probability of D with C1 is P(D and C1) = P(C1)*P(D/C1) = 0.2*0.04 =
0.008. Therefore, P(D) = 0.010 + 0.009 + 0.008 = 0.027
Thus, using the Bayes Theorem (which is simply the application of conditional
probability rule discussed above)
P(C1/D) = [P(D/C1)*P(C1)]/P(D) = [0.02*0.5]/0.027 = 0.37.
Before we found that the item was defective, the probability of the first factory was
50%. After we know that the item is defective, it drops to 37% because the first
factory has relatively the lowest rate of producing defective products.

You might also like