Information Technology Reference
In-Depth Information
Fig. 12.1 The overall system
Retroduction
Abstraction
Concepts
Features
Hypotheses
Rank
Controller
Result
Acceptability
Criteria
Heuristic
Induction
Deduction
Frame
Fig. 12.2 Running
probabilities 'with' (a Direct
learning) and 'without' (b
Window learning) event
memory
{
(X 0 ,X 1 ,X 2 , ...., X n )
(X 1 ,X 2 X 3 , ...., X n + 1 ),
(X 2 ,X 3 ,X 4 , ...., X n + 2 )
...
(X n m ,X n m + 1 , ....,X m )
}
A running probability matrix (see b in Fig. 12.2 ) is used to trace the occurrences of
successful hypothesis. A function to update the running probability matrix is given by
Addis ( 1985 , p. 260) and is also used for belief adjustment as described in Chap. 11.
So we have:
nP(H) t =
(n-1) P(H) t-1
+ α
 
Search WWH ::




Custom Search