Homework Week 10
Homework Week 10
Exercises
1. Given is a T -period dynamic program {S, A, T, (rt )t=1 , (ft )t=1 , (t )t=1 }, where s1 = W ( for some W > 0 );
T T T
( t = 1, . . . , T ); ( t = 1, . . . , T ); s}, ( t = 1, . . . , T ).
Determine an optimal strategy for this program. 2. Given is a T -period dynamic program {S, A, T, (rt )t=1 , (ft )t=1 , (t )t=1 }, where s1 = W ( for some W > 0 );
T T T
Determine an optimal strategy for this program. 3. A friend of yours has W pounds to spend (W > 0) over T periods. He cannot spend more than he has. And whenever he spends something, that money is gone and cannot be used again. He also does not earn any extra money. Finally, if in any period he has s pounds in his pocket and spends a of it, then his utility in that period is u(s, a) = a (2 s a). (a) Write the utility maximisation problem of your friend as a nite-horizon dynamic program, describing the state space S, the action space A, the reward functions rt , etc. 1
Your friend cannot be bothered with all this complicated stu, so he just follows the following strategy: I just spend some of what I have whatever I feel like, without thinking too much about the future; except for the nal period T in which I spend everything left. (b) Show that your friends strategy is in fact always an optimal strategy! 4. Reading: Book by Sundaram: As indicated in Notes 8.