Reasoning with uncertain data [8]

Example logical expressions and CF values to illustrate how CFs are combined:
ExampleMethodResultJustification
IF a AND b
CFa=60% CFb=80%
MinimumPremise is true, CF=60% Both a and b must be true for the premise to be true: the overall certainty is driven by the weakest certainty.
IF a OR b
CFa=60% CFb=80%
MaximumPremise is true, CF=80% The premise is true if either a or b is true: the overall certainty is driven by the strongest certainty.
IF a AND b
IF a OR b
CFa=60% CFb=80%
AveragePremise is true, CF=70% A compromise between the Maximum and Minimum methods for either premise.
CFa=70% from 1st source
CFa=60% from 2nd source
Probability SumFinal CFa=88%
70+(60/100)x(100-70)
30% uncertainty remains from the first certainty factor of 70%. The second CF explains 60% of this remaining 30%. (Note: applying the two certainty values in reverse order yields the same result.)
IF a AND b
Then c=5 with 80% confidence
(premise CF=90%)
Multiplicationc is 5 with 72% confidence The confidence assigned to the attribute in the rule conclusion is assumed independent of the premise confidence level, so the two are combined by multiplying as independent probabilities would be combined.
Expert system software either implements default methods for combining CFs or provides the knowledge base developer with the option to specify methods.

[Next] [Previous] [1 2 3 4 5 6 7 8 9 10 [Home]

Inference Methods and Uncertainty Copyright © 2001-2009 by eXpertise2Go.com. All rights reserved.