BayesTheoremx

Download Report

Transcript BayesTheoremx

An Intuitive Explanation of Bayes'
Theorem
By Eliezer Yudkowsky
reference

http://yudkowsky.net/rational/bayes
Questions

100 out of 10,000 women at age forty who participate in
routine screening have breast cancer. 80 of every 100
women with breast cancer will get a positive
mammography. 950 out of 9,900 women without breast
cancer will also get a positive mammography. If 10,000
women in this age group undergo a routine screening,
about what fraction of women with positive
mammographies will actually have breast cancer?
Compute







A= positive mammographies & actually have breast cancer
B = positive mammographies
results = A/B *100%
A=100*(80/100)=80
B= women with breast cancer with a positive mammography (A) +
women without breast cancer with a positive mammography (C)
B= 100 * (80/100) +(10000-100) *(950/9900)
results = A/B *100% =80/1030=7.8%
Questions – different version

1% of women at age forty who participate in routine
screening have breast cancer. 80% of women with breast
cancer will get positive mammographies. 9.6% of women
without breast cancer will also get positive
mammographies. A woman in this age group had a
positive mammography in a routine screening. What is
the probability that she actually has breast cancer?

A=1%*80%
B=A+(1-1%)*9.6%
Results=A/B*100%=7.8%


Egg problem

Some eggs are painted red and some are painted blue. 40% of the eggs in
the bin contain pearls, and 60% contain nothing. 30% of eggs containing
pearls are painted blue, and 10% of eggs containing nothing are painted
blue. What is the probability that a blue egg contains a pearl?

p(pearl) = 40%
p(blue|pearl) = 30%
p(blue|~pearl) = 10%
p(pearl|blue) = ?





"~" is shorthand for "not", so ~pearl reads "not pearl".
blue|pearl is shorthand for "blue given pearl“


"the probability that an egg is painted blue, given that the egg contains a pearl".
the order of implication is read right-to-left


blue|pearl means "blue<-pearl", the degree to which pearl-ness implies blue-ness
<d|c><c|b><b|a> reads as "the probability that a particle at A goes to B, then to
C, ending up at D".
Visualized Results
More notation



The item on the right side = what you already know or
the premise,
The item on the left side = the implication or conclusion.
p(blue|pearl) = 30%,


p(pearl|blue)



we already know that some egg contains a pearl, then we
can conclude there is a 30% chance that the egg is painted blue.
"the chance that a blue egg contains a pearl" or
"the probability that an egg contains a pearl, if we know the egg is
painted blue"
p(pearl|blue) = p(pear&blue) / p(blue)
Bayes' Theorem

A=1%*80%


B=A+(1-1%)*9.6%


Results=A/B*100%=7.8%

A=p(cancer)*p(positive|cancer)
B=A+ p(~cancer) *p(positive|~cancer)
A/B= p(cancer|positive)
Bayes' Theorem

What we know



P(A)=15%
P(E| A =10%)
What we also know


P(~A)=1- 15% =85%
P(E|~ A) =80%

A

What we want to know

E


Why not 1- 10% ?
Probability of area ? in given E
P(A|E)=?
How?
p(A
| E) =
p(E | A )p(A
)
p(E)
=
p(E | A )p(A
)
p(E | A)p(A ) +p(E |~A)p(~A )
Bayes' Theorem
A1
A2
A3
A4
E
A6
p(A
i
| E) =
A5
p(E | A )p(A
i
p(E)
i
)
=
p(E | A )p(A
 p(E
i
| A )p(A
j
j
where {Ai} forms a partition of the event space,
•Based on definition of conditional probability
•p(Ai|E) is posterior probability given evidence E
•p(Ai) is the prior probability
•P(E|Ai) is the likelihood of the evidence given Ai
•p(E) is the preposterior probability of the evidence
i
)
j
)
p(A
i
| E) =
p(E | A )p(A
i
p(E)
likelihood*prior
Posterior= evidence
i
)
Example

You go to the doctor’s office, where you take a test for a
horrible disease

The test is 99% accurate



If the test is positive, 99% of the time if you have the disease, negative
99% if you don’t
The disease itself is rare: occurs in I in 10,000 people
Your test is positive. What is the probability you have the
disease?
Why is this useful?




14
Useful for assessing diagnostic probability from causal
probability
P(cause|effect)= P(effect|cause)P(cause)
P(effect)
Let M be meningitis, S be stiff neck
P(m|s)=P(s|m)P(m) = 0.8 X 0.0001 = 0.0008
P(s)
0.1
Note: posterior probability of meningitis is still very
small!
Homework


Write a simulation for Monty Hall problem
Suppose you're on a game show, and you're given the
choice of three doors: Behind one door is a car; behind
the others, goats.You pick a door, say No. 1, and the host,
who knows what's behind the doors, opens another door,
say No. 3, which has a goat. He then says to you, "Do you
want to pick door No. 2?" Is it to your advantage to
switch your choice?