Scribd Logo
Upload

Welcome to Scribd!

  • 49 views

    Baye's THM - Lecture 12 Notes

    Uploaded by

    Raiyanul Islam Siam
    The document provides lecture notes on Bayes' theorem that: 1) Defines conditional probability and derives Bayes' theorem from equations for conditional probability. 2) Explains that Bayes' theorem provides a method for updating or revising a probability given additional evidence or information. 3) Provides examples of applying Bayes' theorem to calculate posterior probabilities given prior probabilities and likelihood terms.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Baye's THM - Lecture 12 Notes

    Uploaded by

    Raiyanul Islam Siam
    49 views5 pages
    The document provides lecture notes on Bayes' theorem that: 1) Defines conditional probability and derives Bayes' theorem from equations for conditional probability. 2) Explains that Bayes' theorem provides a method for updating or revising a probability given additional evidence or information. 3) Provides examples of applying Bayes' theorem to calculate posterior probabilities given prior probabilities and likelihood terms.

    Original Title

    Baye's Thm_Lecture 12 Notes

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document provides lecture notes on Bayes' theorem that: 1) Defines conditional probability and derives Bayes' theorem from equations for conditional probability. 2) Explains that Bayes' theorem provides a method for updating or revising a probability given additional evidence or information. 3) Provides examples of applying Bayes' theorem to calculate posterior probabilities given prior probabilities and likelihood terms.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    49 views5 pages

    Baye's THM - Lecture 12 Notes

    Uploaded by

    Raiyanul Islam Siam
    The document provides lecture notes on Bayes' theorem that: 1) Defines conditional probability and derives Bayes' theorem from equations for conditional probability. 2) Explains that Bayes' theorem provides a method for updating or revising a probability given additional evidence or information. 3) Provides examples of applying Bayes' theorem to calculate posterior probabilities given prior probabilities and likelihood terms.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 5

    Bayes’ Theorem Lecture Notes

    Conditional Probability (Review):

    Conditional probability is the probability of one event occurring based on the occurrence of a
    previous event.

    From the diagram above we get two equations:


    𝑃(𝐴⋂𝐵)
    • 𝑃(𝐴|𝐵) = (1.1)
    𝑃(𝐵)

    𝑃(𝐵⋂𝐴)
    • 𝑃(𝐵|𝐴) = 𝑃(𝐴)
    (1.2)

    Rearranging equation 1.1 we get:

    𝑃(𝐴⋂𝐵) = 𝑃(𝐴|𝐵) ∗ 𝑃(𝐵) (1.3)

    Rearranging equation 1.2 we get:

    𝑃(𝐵⋂𝐴) = 𝑃(𝐵|𝐴) ∗ 𝑃(𝐴) (1.4)

    The LHS of both equations 1.3 and 1.4 are equal. Equating them gives:

    𝑃(𝐴|𝐵) ∗ 𝑃(𝐵) = 𝑃(𝐵|𝐴) ∗ 𝑃(𝐴) (1.5)


    𝑃(𝐵|𝐴)∗𝑃(𝐴)
    𝑃(𝐴|𝐵) = 𝑃(𝐵)
    (1.6)

    Equation 1.6 represents the equation of Bayes’ Theorem.

    • Bayes’ Theorem provides a mathematical rule for revising an estimate or forecast in


    light of experiences and observations.
    • Suppose we have estimated prior probabilities for events, and then we obtain new
    information/evidence. Bayes’ Theorem gives us a way to determine the updated or
    posterior probability based on the new evidence.
    Bayes’ Theorem is a method of revising a probability given that additional information is
    obtained. For two events A and B, the Bayes’ Theorem is as follows:

    • Posterior Probability: A revised probability based on additional information


    • Prior Probability: Initial probability based on the present level of information
    • Likelihood Term
    • Evidence

    Total Probability (Review):

    • Let the rectangle be a sample space


    • A1, A2, A3 and A4 are mutually exclusive and collectively exhaustive events
    • Let B be another event in the sample space

    𝑃(𝐵) = 𝑃(𝐵 ⋂𝐴1 ) + 𝑃(𝐵⋂𝐴2 ) + 𝑃(𝐵⋂𝐴3 ) + 𝑃(𝐵⋂𝐴4 )

    = 𝑃(𝐵|𝐴1 ) ∗ 𝑃(𝐴1 ) + 𝑃(𝐵|𝐴2 ) ∗ 𝑃(𝐴2 ) + 𝑃(𝐵|𝐴3 ) ∗ 𝑃(𝐴3 ) + 𝑃(𝐵|𝐴4 ) ∗ 𝑃(𝐴4 )


    4

    = ∑ 𝑃(𝐵|𝐴𝑖 ) ∗ 𝑃(𝐴𝑖 )
    𝑖=1

    In general: 𝑷(𝑩) = ∑𝒌𝒊 𝑷(𝑩|𝑨𝒊 ) ∗ 𝑷(𝑨𝒊 )

    If there are three events A1, A2 & B where A1 & A2 are mutually exclusive and collectively
    exhaustive events then the workable form of Bayes’ theorem is,
    𝑷(𝑩|𝑨𝟏 ) ∗ 𝑷(𝑨𝟏 ) 𝑷(𝑩|𝑨𝟏 ) ∗ 𝑷(𝑨𝟏 )
    𝑷(𝑨𝟏 |𝑩) = =
    𝑷(𝑩) 𝑷(𝑩|𝑨𝟏 ) ∗ 𝑷(𝑨𝟏 ) + 𝑷(𝑩|𝑨𝟐 ) ∗ 𝑷(𝑨𝟐 )

    Thus, for n events A1, A2, ...., Ai, ..., An, the form of Bayes’ theorem is,
    𝑷(𝑩|𝑨𝒊 ) ∗ 𝑷(𝑨𝒊 ) 𝑷(𝑩|𝑨𝒊 ) ∗ 𝑷(𝑨𝒊 )
    𝑷(𝑨𝒊 |𝑩) = =
    𝑷(𝑩) 𝑷(𝑩|𝑨𝟏 ) ∗ 𝑷(𝑨𝟏 ) + 𝑷(𝑩|𝑨𝟐 ) ∗ 𝑷(𝑨𝟐 ) + ⋯ 𝑷(𝑩|𝑨𝒏)𝑷(𝑨𝒏 )

    ---------------------------------------------------------------------------------------------------------------------------
    Examples

    1. A consulting firm submitted a bid for a large consulting contract. The firm's management
    felt it had a 50-50 chance of landing the project. However, the agency to which the bid was
    submitted subsequently asked for additional information. Past experiences indicate that
    for 75% of successful bids and 40% of unsuccessful bids the agency asked for additional
    information.
    a) What is the prior probability of the bid being successful (that is, prior to the request
    for additional information)?
    b) What is the conditional probability of a request for additional information given that the bid
    will be ultimately successful?
    c) Compute the posterior probability that the bid will be successful given a request
    for additional information.
    Answer
    Let A1 = the bid is successful
    Let A2 = the bid is unsuccessful
    Let B = additional information requested

    P(A1) = 0.5
    P(A2) = 0.5
    P(B|A1) = 0.75
    P(B|A2) = 0.4

    We can represent this information using a tree diagram:

    P(B⋂A1)=P(B|A1)*P(A1)=0.5*0.75=0.375

    P(B’⋂A1)=P(B’|A1)*P(A1)=0.5*0.25=0.125
    5

    P(B⋂A2)=P(B|A2)*P(A2)=0.5*0.4=0.2

    P(B’⋂A2)=P(B’|A2)*P(A2)=0.5*0.6=0.3

    a) P(A1) = 0.5
    b) P(B|A1) = 0.75
    𝑃(𝐵|𝐴1 )𝑃(𝐴1 ) 𝑃(𝐵|𝐴1 )𝑃(𝐴1 ) 𝑃(𝐵|𝐴1 )𝑃(𝐴1 )
    c) P(A1|B) = = ∑ 𝑃(𝐵|𝐴𝑖)∗𝑃(𝐴𝑖) = 𝑃(𝐵|𝐴
    𝑃(𝐵) 1 )∗𝑃(𝐴1 )+𝑃(𝐵|𝐴2 )∗𝑃(𝐴2 )
    0.75∗0.5
    = (0.75∗0.5)+(0.4∗0.5)
    = 0.652
    ---------------------------------------------------------------------------------------------------------------------------
    2. A particular study showed that 12% of men will likely develop prostate cancer at some
    point in their lives. A man with prostate cancer has 95% chance of a positive test result from
    a medical screening exam. A man without prostate cancer has a 6% chance of getting a
    false positive test result. What is the probability that a man has cancer given he has a
    positive test result?

    Answer
    Let A1 = has prostate cancer
    Let A2 = does not have prostate cancer
    Let B = positive test result

    P(A1) = 0.12
    P(A2) = 0.88
    P(B|A1) = 0.95
    P(B|A2) = 0.06

    P(A1|B) =?

    𝑃(𝐵|𝐴1 ) ∗ 𝑃(𝐴1 ) 𝑃(𝐵|𝐴1 ) ∗ 𝑃(𝐴1 ) 0.95 ∗ 0.12


    𝑃(𝐴1 |𝐵) = = = = 0.683
    𝑃(𝐵) ∑ 𝑃(𝐵|𝐴𝑖) ∗ 𝑃(𝐴𝑖) (0.95 ∗ 0.12) + (0.06 ∗ 0.88)

    The probability that a man has cancer given he has a positive test result is 68.3%.
    ---------------------------------------------------------------------------------------------------------------------------

    3. Suppose that an insurance company classifies people into one of the three classes –
    good risks, average risks and bad risks. Their records indicate that the probabilities
    that good, average, and bad risk persons will be involved in an accident over a 1-year span
    are 0.05, 0.15 and 0.3 respectively. 20% of the population are "good risks", 50% are
    "average risks", and 30% are "bad risks”.
    a) What proportion of people have accidents in a fixed year?
    b) If policy holder A had no accidents in 1987, what is the probability that he or she is a good
    risk person?
    Answer
    Let A1 = good risk
    Let A2 = average risk
    Let A3 = bad risk
    Let B = accident occurs

    P(A1) = 0.2
    P(A2) = 0.5
    P(A3) = 0.3
    P(B|A1) = 0.05
    P(B|A2) = 0.15
    P(B|A3) = 0.3

    a) 𝑃(𝐵) = ∑ 𝑃(𝐵|𝐴𝑖)𝑃(𝐴𝑖) = 𝑃(𝐵|𝐴1 ) ∗ 𝑃(𝐴1 ) + 𝑃(𝐵|𝐴2 ) ∗ 𝑃(𝐴2 ) + 𝑃(𝐵|𝐴3 ) ∗ 𝑃(𝐴3 )


    = (0.05 ∗ 0.2) + (0.15 ∗ 0.5) + (0.3 ∗ 0.3) = 0.175
    b) Let B’ = accident does not occur
    P(B’) = 1 – P(B)
    ∴ P(B’) = 1 - 0.175 = 0.825
    and P(B’|A1) = 1 - P(B|A1)
    = 1 - 0.05 = 0.95
    𝑃(𝐵′|𝐴1 ) ∗ 𝑃(𝐴1 ) 0.95 ∗ 0.20
    P(A1 |B′) = = = 0.23
    𝑃(𝐵′) 0.825
    If policy holder A had no accidents in 1987, the probability that he or she is a good risk is
    23%.

    ---------------------------------------------------------------------------------------------------------------------------

    Practice Problems

    Textbook: Probability & Statistics for Engineering and the Sciences (Devore)

    Conditional Probability and Bayes' Theorem

    Page 80-83: 45, 49, 51, 53, 55, 59, 61, 65, 67

    You might also like

    pFad - Phonifier reborn

    Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

    Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


    Alternative Proxies:

    Alternative Proxy

    pFad Proxy

    pFad v3 Proxy

    pFad v4 Proxy