This preview shows page 1. Sign up to view the full content.
8
The dribble success probability
y
in fact depend on the presence or absence of a defending robot,
D
.
A
has
no way of detecting whether
D
is present, but does know some statistical properties of its environment.
D
is present
2
3
of the time. When
D
is absent,
y
=
3
4
. When
D
is present,
y
=
1
4
.
(f) (4 pt)
What is the posterior probability that
D
is present, given that
A
D
ribbles twice successfully from
1 to 3, then
S
hoots from state 3 and scores.
We can use Bayes’ rule, where
D
is a random variable denoting the presence of
D
, and
e
is the evidence
that
A
dribbled twice and scored.
P
(
d

e
) =
P
(
e

d
)
·
P
(
d
)
P
(
e
)
P
(
e
) =
P
(
e

d
)
·
P
(
d
) +
P
(
e
¬
d
)
·
P
(
¬
d
)
P
(
e

d
) =
1
4
·
1
4
·
1
2
P
(
e
¬
d
) =
3
4
·
3
4
·
1
2
P
(
e
) =
1
32
·
2
3
+
9
32
·
1
3
=
11
96
P
(
d

e
) =
2
96
/
11
96
=
2
11
(g) (3 pt)
What transition model should
A
use in order to correctly compute its maximum expected reward
when it doesn’t know whether or not
D
is present?
To maximize expected total reward, the agent should model the situation as accurately as possible.
This is the end of the preview. Sign up
to
access the rest of the document.
This note was uploaded on 08/30/2009 for the course CS 188 taught by Professor Staff during the Spring '08 term at University of California, Berkeley.
 Spring '08
 Staff
 Artificial Intelligence

Click to edit the document details