Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
All balls in the box are black This ball is from the box This ball is black All balls in the box are black This ball is black This ball is from the box These balls are from the box These balls are black All ball in the box are black
A => B A --------B A => B B ------------Possibly A Whenever A then B but not vice versa ------------Possibly A => B
Default reasoning : Assume x to be true when nothing is known to the contrary. I.e. no evidence is now available that says ~x
2
Given Doorbell, can we say AtDoor(x), because AtDoor(x) Doorbell? Abductive Reasoning But No, the doorbell might start ringing due to some other reason Eg
Short circuit Wind animals
Given Doorbell, can we say wake(Mohan), because Doorbell Wake(Mohan)? Deductive Reasoning Yes, only IF Proposition 2 is always true. However, in general Mohan may not always wake up, even if the bell rings.
Therefore we cannot answer either of the questions with certainty. Proposition 1 is incomplete. Modifying it as
AtDoor(x) V ShortCkt V Wind . Doorbell Doesn t help because the list of possible causes on
However, problem like that of the doorbell are very common in real life In AI, we often need to reason under such circumstances
We solve it by proper modelling of uncertainty and impreciseness, and developing appropriate reasoning techniques.
frequencies
Doorbell
Incomplete knowledge
We may not know or guess all the possible
Conflicting information
Expert often provide conflicting information Quantification of measure of belief
Propagation of uncertainties
In absence of interdependencies propagation of
uncertain knowledge increases the uncertainty of the conclusions Tomorrow(sunny)[6.0], Tomorrow(warm) [8.0] Tomorrow(sunny) Tomorrow(warm) [?]
to occur
Non-negative value in [0,1] Total probability of the sample space is 1
What is the probability that my house was robbed today? I think there s about 10 percent chance I was robbed today
Expresses the degree to which I believe this claim
to be true
reasoning from bad reasoning How can a subjective theory of probability do this?
$ 1000
1000 ticket
Bet I :
P(head ) = 0.50
P(Win) = 0.25
Step2: Show that a rational betting strategy must satisfy the rules of probability theory
No rational person will willingly agree to a bet
For mutually exclusive events, the probability for at least one of them is the sum of their individual probabilities Experimental probability
Based on the frequency of events
Subjective probability
Based on expert assessment
A value between 0 and 1. P(E) denotes the probability of the event E. P(True) = 1 , i.e. certain event P(False) = 0 , i.e. impossible event P(A) = 1 P(~A) P(AUB) = P(A) + P(B) P(A B)
Conditional probability
Of event a given that event B has already occurred P(A|B) = P(A B) / P(B)