1 / 7

Reasoning with Uncertainty

Reasoning with Uncertainty. We have accepted that (A -> B) is TRUE followed by A is TRUE, would result in B is TRUE (that is, modus ponens rule is true) e.g. A := it rains B := umbrella opens

starbuck
Download Presentation

Reasoning with Uncertainty

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Reasoning with Uncertainty • We have accepted that (A -> B) is TRUE followed by A is TRUE, would result in B is TRUE • (that is, modus ponens rule is true) • e.g. • A := it rains • B := umbrella opens • 1) A -> B 1) A -> B • 2) A 2) B • -------------------- ----------------- • 3) B 3) A ? • What about the reverse? : A -> B, followed by B is true • (Does that mean that “it must be raining?”)

  2. Not Necessarily (that it is raining) A B A -> B A -> B is true and B is true does that mean A is true? TTT T F F FTT Not necessarily because the umbrella may be open to protect us from bright sun also ! F F T

  3. Uncertainty with Non-Independent Eventsand “Probabilistic” Logic Consider the following (weather and behavior) events: - if it rains then umbrella opens : A -> B is TRUE; - there is a 30% chance of (raining Λ umbrella opens); - there is a 5% chance of (sunny Λ and umbrella opens) .30 .70 .35 .65 If the umbrella is open does that mean that it is raining? Answer is “may be” because: P (raining | umbrella is open) = P (raining Λ umbrella open) / P (umbrella open) = .30 / .35 = 6/7 chance that it is raining when umbrella is open

  4. How About the Original Case? If it is raining, does that mean the umbrella is open? Answer: “definitely” P (umbrella is open | raining) = P (raining Λ umbrella open) / P (raining) = .30 / .30 = 1/1 = 100% chance that umbrella is open if it rains Thus : A -> B A ----------------- B is 100% and thus modus ponens works as a syllogism

  5. What about other conditions? If the umbrella is open, we said that there is a 6/7 or 86% chance that it is raining. However, if the umbrella is open there is also a 1/7 or 14% chance that it is not raining! - P( not raining| umbrella open) = P (not raining Λ umbrella open)/P (umbrella open) = .05 / .35 = 1/7 = 14% What about modus tollens? A -> B ~B ----------------- ~ A P(~A | ~B) = P (~ A Λ ~B) / P ( ~B) = .65 /.65 = 100%

  6. Interesting Scenarios From the sample - “ -if it is raining the umbrella will be open; -it is raining; therefore, the umbrella is open” we actually have some of the following interesting cases: 1) P( umbrella is open | raining) = .30/.30 = 100% (modus ponens) 2) P( raining | umbrella is open) = .30/.35 = 6/7 = 86% 3) P( not raining | umbrella is open) = .05/.35 = 14% 4) P( not raining | umbrella is not open) = .65/.65 = 100% (modus tollens)

  7. Look at “simple pseudo-code” of non-independent events Pseudo-code segment: Let A : k = true and B: y = 80 & our code segment shows A -> B . . if (k is true) { z := 50 ; y := z +30; } else { y := z + 50; } print y . . Question: if “print y” shows y = 80, is k = true? Answer: - most likely but NOT necessarily: P(A | B) = P( A Λ B) / P(B) - if P(A Λ B) = P(B), then yes. - but we must check the code “above and below” the shown code segment (e.g. is Z set to 30 before? ) Question: print y shows y ≠ 80, is k = true? Answer: - NO (via modus tollens)

More Related