Easy to understand Bayes formula
To understand Bayes formula , First, we need to know the conditional probability formula , The concept of total probability formula , Because the Bayes formula is derived from both of them .
One , Conditional probability formula
P(B|A)= P(AB)/(P(A)
meaning ： condition B Lower generation A The probability of A and B The joint probability divided by B The probability of .
Total probability formula , Bayes formula is derived from it .
P(AB)
It's an event A happen , And events B Probability of occurrence , Event scope or entire sample space ;A,B There is no order , Among them AB mean AB The part where two spaces intersect , This part can also be regarded as in the B In space ,A And there it is , In fact, this formula P(AB) So Ω Of the sample space , however P(A|B) So B Of the sample space ,
P(B | A) It's an event A First ,B Probability of occurrence , in other words P(B | A) Greater than or equal to P(AB). Be aware that it's an event A First , Because of events A The sample space is narrowed by the occurrence of .
P(AB)：
P(B | A)：
Two , Multiplication theorem
P(AB) = P( B | A)P(A), P(AB) = P( A | B)P(B)
It's easy to understand with the above pictures , event A And events B The probability of simultaneous occurrence is equal to the event B In the event A The probability of occurrence times the event in the sample space A The probability of occurrence in the entire sample space .
therefore , The rudiment of Bayes formula
From the above two formulas ：P(B|A)= (P(A|B)P(B))/(P(A))
Three , Total probability formula
set up E For the experiment A Sample space of ,B1,B2…Bn by A A set of events for , Namely B The series of events is divided up A And each B No overlap （ Incompatible ）, Then the total probability formula can be deduced as ：
When B1+B2+B3=A Time ：
P(A)
= P(AA)
=P[A(B1+B2+B3)]
=P(AB1+AB2+AB3)
=P(AB1)+P(AB2)+P(AB3),（ because P(AB)=P(A|B)P(B)）
=P(A|B1)P(B1)+ P(A|B2)P(B2) + P(A|B3)P(B3)
Namely ：P(A) = P(B1) P(A|B1) + P(B2)P(A|B2) + …
Four , Bayes formula ：
It can be obtained from the formula of conditional probability and total probability ：
Technology