Paper 2, Section II, K

Optimization and Control
Part II, 2011

Consider an optimal stopping problem in which the optimality equation takes the form

Ft(x)=max{r(x),E[Ft+1(xt+1)]},t=1,,N1,F_{t}(x)=\max \left\{r(x), E\left[F_{t+1}\left(x_{t+1}\right)\right]\right\}, \quad t=1, \ldots, N-1,

FN(x)=r(x)F_{N}(x)=r(x), and where r(x)>0r(x)>0 for all xx. Let SS denote the stopping set of the onestep-look-ahead rule. Show that if SS is closed (in a sense you should explain) then the one-step-look-ahead rule is optimal.

NN biased coins are to be tossed successively. The probability that the ii th coin toss will show a head is known to be pi(0<pi<1)p_{i}\left(0<p_{i}<1\right). At most once, after observing a head, and before tossing the next coin, you may guess that you have just seen the last head (i.e. that all subsequent tosses will show tails). If your guess turns out to be correct then you win £1£ 1.

Suppose that you have not yet guessed 'last head', and the ii th toss is a head. Show that it cannot be optimal to guess that this is the last head if

pi+1qi+1++pNqN>1\frac{p_{i+1}}{q_{i+1}}+\cdots+\frac{p_{N}}{q_{N}}>1

where qj=1pjq_{j}=1-p_{j}.

Suppose that pi=1/ip_{i}=1 / i. Show that it is optimal to guess that the last head is the first head (if any) to occur after having tossed at least ii^{*} coins, where iN/ei^{*} \approx N / e when NN is large.