Please enable JavaScript.
Coggle requires JavaScript to display documents.
Ch4 662 : Uncertainty in Rule Based (Certainty Factor (Unknown (-0.2 to +0…
Ch4 662 : Uncertainty in Rule Based
Source of uncertain Knowledge
Weak implication
Domain expert has a
difficult
task to establish
concrete correlation
between
IF-THEN rules
Imprecise language
We describe facts with
often
,
frequently
or
hardly ever
. It can be
difficult to express
that knowledge in IF-THEN rules
Unknown Data
When
data is incomplete
, the only solution is to
accept the "unknown" data
and
proceed to predict reasoning
with this value
Combining the View of Different Expert
Multiple experts
will cause
different opinion
and cause
conflicting rules
Uncertainty in AI
Information is partial
Information is not fully reliable
Conflicting information
Information is approximate
Types of error
Ambiguity
Something may be interpreted in more than 1 way
Incomplete
Some information is missing
Incorrect
The information is wrong
Measurement
Error in precision and accuracy
Unreliability
If the measuring equipment is unreliable, the data will have error
Random Error
Caused by unknown factor. It will lead to uncertainty of the mean
Systematic Error
Not random but instead it is introduced because some bias
Certainty Factor
Unknown
-0.2 to +0.2
Maybe
+0.4
Maybe Not
-0.4
Probably Not
-0.6
Probably
+0.6
Almost certainly not
-0.8
Almost certainly
+0.8
Definitely not
-1.0
Definitely
+1.0
Uncertainty
Defined as the
lack of exact knowledge
that would enable us to reach a perfectly reliable conclusion
Calculation
Kena tahu cara caculate
Certainty Factor (CF)
Hafal formula dua2 positive, positive&negative, dua2 negative
Bayes Theorem
Hafal formula probability yg
A given B
bagai