Information Technology Reference
In-Depth Information
The state H is the quality of the movie, the possible values of
which are members of the set V H
(but
the prevailing state s for the current game is to be determined).
=
{
Good , Average , Bad
}
( I i ) is as shown in Tables 3.2 to 3.5. For example, agent A 1 's ex-
perience I 1 as shown in Table 3.2 has 9 rules, which are
(Director = Spielberg)
(Type = S.F.)
Good
(Director = Spielberg)
(Type = Biographic)
Good
(Director = Spielberg) (Type = S.F.) Good
(Director = Coppola)
(Type = Drama)
Good
(Director = Tempa)
(Type = Action)
Bad
(Director = Emmerich)
(Type = S.F.)
Bad
(Director = Sonnenfield)
(Type = S.F.)
Good
(Director = Spielberg)
(Type = S.F.)
Good
(Director = Spielberg)
(Type = Drama)
Average
Some of the possible consequences for this coalition are (tennis, ten-
nis, tennis, tennis), (tennis, tennis, movie, movie), (movie, movie,
movie, movie) . . . .
P i is agent i 's set of preference rules. Intuitively, each rule de-
termines part of the agent i 's preference, given the states in the
agent's experience I i that supported by the current evidences E .
For example, as currently
E = { (Director = Spielberg) (Type = S.F.) },
by agent A 1 's experience I 1 the supported states are members of
the set
{
Good , Good , Good
}
=
{
Good
}
.
As we assume that 'if the movie is good, then all four agents prefer
the movie over tennis,' we would expect that P 1 contains, among
Search WWH ::




Custom Search