Skip to Main Content
Automated Planning
book

Automated Planning

by Malik Ghallab, Dana Nau, Paolo Traverso
May 2004
Intermediate to advanced content levelIntermediate to advanced
635 pages
19h 46m
English
Morgan Kaufmann
Content preview from Automated Planning
390 Chapter 16 Planning Based on Markov Decision Processes
Value-Iteration (,C,γ )
for each s S, select any initial value E
0
(s)
k 1
while k<maximum number of iterations do
for each s S do
for each a A do
Q(s, a) C(s, a) + γ
s
S
P
a
(s
|s) E
k1
(s
)
E
k
(s) min
aA
Q(s, a)
π(s) arg min
aA
Q(s, a)
k k + 1
return(π)
end
Figure 16.6 Value iteration.
is the following:
max
sS
|E
n
(s) E
n1
(s)| < (16.11)
This stopping criterion guarantees that the returned policy is an -optimal policy,
i.e., it has an expected cost that does not differ from the optimum by more than an
arbitrarily small number .
Example 16.8 Consider again the situation shown in Figure 16.5. ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Start your free trial

You might also like

Communicate with Teams More Effectively

Communicate with Teams More Effectively

Charles Humble
How to Overcome a Power Deficit

How to Overcome a Power Deficit

Cyril Bouquet, Jean-Louis Barsoux

Publisher Resources

ISBN: 9781558608566