| | |
Summary: ISyE 6664 Exam # 1
Fall 2005
Name
Please be neat and show all your work so that I can give you partial credit.
GOOD LUCK.
Question 1
Question 2
Question 3
Question 4
Total
1
1. (25) Consider a model with T = {1, 2}, S = {s1, s2}, As1 = {a1,1, a1,2}
and As2 = {a2,1, a2,2}, r1(s1, a1,1) = 5, r1(s1, a1,2) = 10, r1(s2, a2,1) = -1, and
r1(s2, a2,2) = 2 and p1(s1|s1, a1,1) = p1(s2|s1, a1,1) = 0.5, p1(s1|s1, a1,2) = 0,
p1(s2|s1, a1,2) = 1, p1(s1|s2, a2,1) = 0.8, p1(s2|s2, a2,1) = 0.2, p1(s1|s2, a2,2) =
0.1 and p1(s2|s2, a2,2) = 0.9.
a. (10) Find the deterministic policy that maximizes the total expected reward
provided that the terminal reward for both states is 0.
b. (15) Find the deterministic policy that maximizes the total expected reward
provided that the terminal reward r2(s1) = d and r2(s2) = e. Investigate the
|
