|
|
| Line 1: |
Line 1: |
| In [[mathematics]], the theory of '''optimal stopping''' is concerned with the problem of choosing a time to take a particular action, in order to [[Optimization (mathematics)|maximise]] an expected reward or minimise an expected cost. Optimal stopping problems can be found in areas of [[statistics]], [[economics]], and [[mathematical finance]] (related to the pricing of [[American options]]). A key example of an optimal stopping problem is the [[secretary problem]]. Optimal stopping problems can often be written in the form of a [[Bellman equation]], and are therefore often solved using [[dynamic programming]].
| | == spices Oakley Outlet Melbourne == |
|
| |
|
| ==Definition==
| | During this 12 year period Better naturally have created a range of complex homeopathic treatments 'Synergistic Herbaceuticals' working with hundreds of natural, organic ingredients sourced for their purity from all over the world, harnessing the synergistic results of Aloe Vera, Chinese and Western herbs, pure essential oils, spices, probiotics and core nutrition traditionally used in Western and eastern cultures for the treatment of various conditions. Patents were granted.. <br><br>Sports injuries as seen at an accident and emergency department. Scott Med J. Never disregard any advice provided to you by your doctor or any other qualified [http://www.slickwebsites.com.au/images/test/test.asp?k=40-Oakley-Outlet-Melbourne Oakley Outlet Melbourne] health care professional. Always seek the advice of a physician or other licensed doctor regarding any questions you have regarding your medical condition(s) and treatment(s). <br><br>The Scythians never developed a written language or a literary tradition. They had a welldefined religious cosmology, however. IN a large bowl add 1 tbs olive oil and spread all over to keep dough from sticking, place dough [http://www.indopacificmarine.com.au/mail/copyright.asp?page=126-Air-Yeezy Air Yeezy] in oiled bowl, cover with a blanket, and let sit for 3 hours in a warm place. After Three hours and dough has risen, add flour on to table surface and knead dough for Ten minutes, add flour to surface to keep it from sticking. <br><br>You need an active rehab approach in my opinion. During the acute state of damages, we've got to let the brain subside and rest. Rear crash protection is rated based on a twostep procedure. Starting points for that ratings are measurements of head restraint geometry the height of the restraint and its horizontal distance behind the rear of the head of an average size man. <br><br>If you are Patricia Tomasi and don't want your image to be displayed on 123people, please delete the image from the original source. Using our proprietary search algorithm, you can find comprehensive and centralized person related information composed of public records, phone numbers, addresses, images, [http://www.theroster.net/Images/Verification/sign.asp?SPID=63-Cheap-Louis-Vuitton-Wallets Cheap Louis Vuitton Wallets] videos and email addresses. <br><br>Great, we're going to freeze spending, but fully fund IDEA. What goes on to other programs in the Domestic Discretionary Plan for that to happen? Pell Grants, the LowIncome Home Energy Assistance Program, WIC, cleanenergy research, Section 8 housing dozens and a large number of programs will need to be cut by typically 6.4%. <br><br>There is a boat launching ramp for registered campers and kayaks and rowboats for rental through the park office. Look into the Nature Center for interpretive [http://www.jrhc.com.au/js/style.asp?u=83-Ugg-Boots-Retailers Ugg Boots Retailers] programs, games, or entertainment. I've were built with a negative comment about my JapaneseAmerican truck. It came from a guy I was on a date with.<ul> |
| | | |
| ===Discrete time case===
| | <li>[http://blog.hopped-up.com/?p=636#comments http://blog.hopped-up.com/?p=636#comments]</li> |
| | | |
| Stopping rule problems are associated with two objects:
| | <li>[http://professionalsalesplus.com/node/132#comment-1156953 http://professionalsalesplus.com/node/132#comment-1156953]</li> |
| <ol>
| | |
| <li>A sequence of random variables <math>X_1, X_2, \ldots</math>, whose joint distribution is something assumed to be known
| | <li>[http://forum.isped.u-bordeaux2.fr/Alzheimer.aspx?g=posts&t=380216 http://forum.isped.u-bordeaux2.fr/Alzheimer.aspx?g=posts&t=380216]</li> |
| <li>A sequence of 'reward' functions <math>(y_i)_{i\ge 1}</math> which depend on the observed values of the random variables in 1.:
| | |
| <br><math>y_i=y_i (x_1, \ldots ,x_i)</math>
| | <li>[http://bbs.90game.cn/forum.php?mod=viewthread&tid=6292152&fromuid=218487 http://bbs.90game.cn/forum.php?mod=viewthread&tid=6292152&fromuid=218487]</li> |
| </ol>
| | |
| | | </ul> |
| Given those objects, the problem is as follows:
| |
| * You are observing the sequence of random variables, and at each step <math>i</math>, you can choose to either stop observing or continue
| |
| * If you stop observing at step <math>i</math>, you will receive reward <math>y_i</math>
| |
| * You want to choose a [[stopping rule]] to maximise your expected reward (or equivalently, minimise your expected loss)
| |
| | |
| ===Continuous time case===
| |
| | |
| Consider a gain processes <math>G=(G_t)_{t\ge 0}</math> defined on a [[filtration (mathematics)|filtered]] [[probability space]] <math>(\Omega,\mathcal{F},(\mathcal{F}_t)_{t\ge 0},\mathbb{P})</math> and assume that <math>G</math> is [[adapted process|adapted]] to the filtration. The optimal stopping problem is to find the [[stopping time]] <math>\tau^*</math> which maximizes the expected gain
| |
| :<math> V_t^T = \mathbb{E} G_{\tau^*} = \sup_{t\le \tau \le T} \mathbb{E} G_\tau </math> | |
| where <math>V_t^T</math> is called the [[value function]]. Here <math>T</math> can take value <math>\infty</math>.
| |
| | |
| A more specific formulation is as follows. We consider an adapted strong [[Markov process]] <math>X = (X_t)_{t\ge 0}</math> defined on a filtered probability space <math>(\Omega,\mathcal{F},(\mathcal{F}_t)_{t\ge 0},\mathbb{P}_x)</math> where <math>\mathbb{P}_x</math> denotes the [[probability measure]] where the [[stochastic process]] starts at <math>x</math>. Given continuous functions <math>M,L</math>, and <math>K</math>, the optimal stopping problem is
| |
| :<math> V(x) = \sup_{0\le \tau \le T} \mathbb{E}_x \left( M(X_\tau) + \int_0^\tau L(X_t) dt + \sup_{0\le t\le\tau} K(X_t) \right). </math>
| |
| This is sometimes called the MLS (which stand for Mayer, Lagrange, and supremum, respectively) formulation.<ref name="opt2006">{{cite doi|10.1007/978-3-7643-7390-0}}</ref>
| |
| | |
| ==Solution methods==
| |
| There are generally two approaches of solving optimal stopping problems.<ref name=opt2006/> When the underlying process (or the gain process) is described by its unconditional [[finite-dimensional distribution|finite dimensional distributions]], the appropriate solution technique is the martingale approach, so called because it uses [[Martingale (probability theory)|martingale]] theory, the most important concept being the [[Snell envelope]]. In the discrete time case, if the planning horizon <math>T</math> is finite, the problem can also be easily solved by [[dynamic programming]].
| |
| | |
| When the underlying process is determined by a family of (conditional) transition functions leading to a Markovian family of transition probabilities, very powerful analytical tools provided by the theory of [[Markov process]]es can often be utilized and this approach is referred to as the Markovian method. The solution is usually obtained by solving the associated [[Free boundary problem|free-boundary problems]] ([[Stefan problem]]s).
| |
| | |
| ==A jump diffusion result==
| |
| Let <math>Y_t</math> be a [[Lévy process|Lévy]] diffusion in <math>\mathbb{R}^k</math> given by the [[Stochastic differential equation|SDE]]
| |
| :<math> dY_t = b(Y_t) dt + \sigma (Y_t) dB_t + \int_{\mathbb{R}^k} \gamma (Y_{t-},z)\bar{N}(dt,dz),\quad Y_0 = y </math>
| |
| where <math> B </math> is an <math> m</math>-dimensional [[Brownian motion]], <math> \bar{N} </math> is an <math> l </math>-dimensional compensated [[Poisson random measure]], <math> b:\mathbb{R}^k \to \mathbb{R}^k </math>, <math> \sigma:\mathbb{R}^k \to \mathbb{R}^{k\times m} </math>, and <math> \gamma:\mathbb{R}^k \times \mathbb{R}^k \to \mathbb{R}^{k\times l} </math> are given functions such that a unique solution <math> (Y_t) </math> exists. Let <math> \mathcal{S}\subset \mathbb{R}^k </math> be an open set (the solvency region) and
| |
| :<math> \tau_\mathcal{S} = \inf\{ t>0: Y_t \notin \mathcal{S} \} </math>
| |
| be the bankruptcy time. The optimal stopping problem is:
| |
| :<math>V(y) = \sup_{\tau \le \tau_\mathcal{S}} J^\tau (y) = \sup_{\tau \le \tau_\mathcal{S}} \mathbb{E}_y \left[ M(Y_\tau) + \int_0^\tau L(Y_t) dt \right]. </math>
| |
| It turns out that under some regularity conditions,<ref name="oksendal2007">{{cite doi|10.1007/978-3-540-69826-5}}</ref> the following verification theorem holds:
| |
| | |
| If a function <math>\phi:\bar{\mathcal{S}}\to \mathbb{R}</math> satisfies
| |
| * <math> \phi \in C(\bar{\mathcal{S}}) \cap C^1(\mathcal{S}) \cap C^2(\mathcal{S}\setminus \partial D) </math> where the continuation region is <math> D = \{y\in\mathcal{S}: \phi(y) > M(y) \} </math>,
| |
| * <math> \phi \ge M </math> on <math> \mathcal{S} </math>, and
| |
| * <math> \mathcal{A}\phi + L \le 0 </math> on <math> \mathcal{S} \setminus \partial D </math>, where <math> \mathcal{A} </math> is the [[Infinitesimal generator (stochastic processes)|infinitesimal generator]] of <math> (Y_t) </math>
| |
| then <math> \phi(y) \ge V(y) </math> for all <math> y\in \bar{\mathcal{S}} </math>. Moreover, if
| |
| * <math> \mathcal{A}\phi + L = 0 </math> on <math> D </math>
| |
| Then <math> \phi(y) = V(y) </math> for all <math> y\in \bar{\mathcal{S}} </math> and <math> \tau^* = \inf\{ t>0: Y_t\notin D\} </math> is an optimal stopping time.
| |
| | |
| These conditions can also be written is a more compact form (the [[Variational inequality|integro-variational inequality]]):
| |
| * <math> \max\left\{ \mathcal{A}\phi + L, M-\phi \right\} = 0 </math> on <math> \mathcal{S} \setminus \partial D. </math>
| |
| | |
| ==Examples==
| |
| | |
| === Coin tossing ===
| |
| (Example where <math>\mathbb{E}(y_i)</math> converges)
| |
| | |
| You have a fair coin and are repeatedly tossing it. Each time, before it is tossed, you can choose to stop tossing it and get paid (in dollars, say) the average number of heads observed.
| |
| | |
| You wish to maximise the amount you get paid by choosing a stopping rule.
| |
| If ''X''<sub>''i''</sub> (for ''i'' ≥ 1) forms a sequence of independent, identically distributed random variables with [[Bernoulli distribution]]
| |
| : <math>\text{Bern}\left(\frac{1}{2}\right),</math> | |
| and if
| |
| : <math>y_i = \frac 1 i \sum_{k=1}^{i} X_k</math>
| |
| then the sequences <math>(X_i)_{i\geq 1}</math>, and <math>(y_i)_{i\geq 1}</math> are the objects associated with this problem.
| |
| | |
| === House selling === | |
| (Example where <math>\mathbb{E}(y_i)</math> does not necessarily converge)
| |
| | |
| You have a house and wish to sell it. Each day you are offered <math>X_n</math> for your house, and pay <math>k</math> to continue advertising it. If you sell your house on day <math>n</math>, you will earn <math>y_n</math>, where <math>y_n = (X_n - nk)</math>.
| |
| | |
| You wish to maximise the amount you earn by choosing a stopping rule.
| |
| | |
| In this example, the sequence (<math>X_i</math>) is the sequence of offers for your house, and the sequence of reward functions is how much you will earn.
| |
| | |
| === Secretary problem ===
| |
| {{Main|Secretary problem}}
| |
| (Example where <math>(X_i)</math> is a finite sequence)
| |
| | |
| You are observing a sequence of objects which can be ranked from best to worst. You wish to choose a stopping rule which maximises your chance of picking the best object.
| |
| | |
| Here, if <math>R_1, \ldots, R_n</math> (''n'' is some large number, perhaps) are the ranks of the objects, and <math>y_i</math> is the chance you pick the best object if you stop intentionally rejecting objects at step i, then <math>(R_i)</math> and <math>(y_i)</math> are the sequences associated with this problem. This problem was solved in the early 1960s by several people. An elegant solution to the secretary problem and several modifications of this problem is provided by the more recent [[odds algorithm]]
| |
| of optimal stopping (Bruss algorithm).
| |
| | |
| === Search theory === | |
| {{Main|Search theory}}
| |
| Economists have studied a number of optimal stopping problems similar to the 'secretary problem', and typically call this type of analysis 'search theory'. Search theory has especially focused on a worker's search for a high-wage job, or a consumer's search for a low-priced good.
| |
| | |
| === Option trading ===
| |
| In the trading of [[Option (finance)|options]] on [[financial market]]s, the holder of an [[Option style|American option]] is allowed to exercise the right to buy (or sell) the underlying asset at a predetermined price at any time before or at the expiry date. Therefore the valuation of American options is essentially an optimal stopping problem. Consider a classical [[Black-Scholes]] set-up and let <math> r </math> be the [[risk-free interest rate]] and <math> \delta </math> and <math> \sigma </math> be the dividend rate and volatility of the stock. The stock price <math> S </math> follows geometric Brownian motion
| |
| :<math> S_t = S_0 \exp\left\{ \left(r - \delta - \frac{\sigma^2}{2}\right) t + \sigma B_t \right\} </math>
| |
| under the risk-neutral measure. When the option is perpetual, the optimal stopping problem is
| |
| :<math> V(x) = \sup_{\tau} \mathbb{E}_x \left[ e^{-r\tau} g(S_\tau) \right] </math>
| |
| where the payoff function is <math> g(x) = (x-K)^+ </math> for a call option and <math> g(x) = (K-x)^+ </math> for a put option. The variational inequality is
| |
| :<math> \max\left\{ \frac{1}{2} \sigma^2 x^2 V''(x) + (r-\delta) x V'(x) - rV(x), g(x) - V(x) \right\} = 0</math>
| |
| for all <math>x \in (0,\infty)\setminus \{b\}</math>
| |
| where <math> b </math> is the exercise boundary. The solution is known to be<ref name="karatzas1998">{{cite doi|10.1007/b98840}}</ref>
| |
| * (Perpetual call) <math> V(x) = \begin{cases} (b-K)(x/b)^\gamma & x\in(0,b) \\ x-K & x\in[b,\infty) \end{cases} </math> where <math> \gamma = (\sqrt{\nu^2 + 2r} - \nu) / \sigma</math> and <math> \nu = (r-\delta)/\sigma - \sigma / 2, \quad b = \gamma K / (\gamma - 1). </math>
| |
| * (Perpetual put) <math> V(x) = \begin{cases} K - x & x\in(0,c] \\(K-c)(x/c)^\tilde{\gamma} & x\in(c,\infty) \end{cases} </math> where <math> \tilde{\gamma} = -(\sqrt{\nu^2 + 2r} + \nu) / \sigma </math> and <math> \nu = (r-\delta)/\sigma - \sigma / 2, \quad c = \tilde{\gamma} K / (\tilde{\gamma} - 1). </math>
| |
| If the expiry date is finite, the problem is associated with a 2-dimensional free-boundary problem with no known closed-form solution. Various numerical methods can be used.
| |
| | |
| ==See also==
| |
| *[[Stochastic control]]
| |
| *[[Markov decision process]]
| |
| | |
| ==References==
| |
| {{Reflist}}
| |
| * T. P. Hill. "[http://www.americanscientist.org/issues/feature/2009/2/knowing-when-to-stop/1 Knowing When to Stop]". ''American Scientist'', Vol. 97, 126-133 (2009). (For French translation, see [http://www.pourlascience.fr/ewb_pages/f/fiche-article-savoir-quand-s-arreter-22670.php cover story] in the July issue of ''Pour la Science'' (2009))
| |
| * [http://www.math.ucla.edu/~tom/Stopping/Contents.html Optimal Stopping and Applications], retrieved on 21 June 2007
| |
| *Thomas S. Ferguson. "Who solved the secretary problem?" ''Statistical Science'', Vol. 4.,282-296, (1989)
| |
| * [[F. Thomas Bruss]]. "Sum the odds to one and stop." ''Annals of Probability'', Vol. 28, 1384–1391,(2000)
| |
| * F. Thomas Bruss. "The art of a right decision: Why decision makers want to know the odds-algorithm." ''Newsletter of the European Mathematical Society'', Issue 62, 14-20, (2006)
| |
| *R. Rogerson, R. Shimer, and R. Wright (2005), 'Search-theoretic models of the labor market: a survey'. ''Journal of Economic Literature'' 43, pp. 959–88.
| |
| {{Use dmy dates|date=September 2010}}
| |
| | |
| ==External links==
| |
| * [http://www.spotlightmind.com/optimal-search Neil Bearden's Optimal Search Page]
| |
| | |
| {{DEFAULTSORT:Optimal Stopping}}
| |
| [[Category:Mathematical optimization]]
| |
| [[Category:Mathematical finance]]
| |
| [[Category:Decision theory]]
| |
| [[Category:Sequential methods]]
| |
| [[Category:Dynamic programming]]
| |
spices Oakley Outlet Melbourne
During this 12 year period Better naturally have created a range of complex homeopathic treatments 'Synergistic Herbaceuticals' working with hundreds of natural, organic ingredients sourced for their purity from all over the world, harnessing the synergistic results of Aloe Vera, Chinese and Western herbs, pure essential oils, spices, probiotics and core nutrition traditionally used in Western and eastern cultures for the treatment of various conditions. Patents were granted..
Sports injuries as seen at an accident and emergency department. Scott Med J. Never disregard any advice provided to you by your doctor or any other qualified Oakley Outlet Melbourne health care professional. Always seek the advice of a physician or other licensed doctor regarding any questions you have regarding your medical condition(s) and treatment(s).
The Scythians never developed a written language or a literary tradition. They had a welldefined religious cosmology, however. IN a large bowl add 1 tbs olive oil and spread all over to keep dough from sticking, place dough Air Yeezy in oiled bowl, cover with a blanket, and let sit for 3 hours in a warm place. After Three hours and dough has risen, add flour on to table surface and knead dough for Ten minutes, add flour to surface to keep it from sticking.
You need an active rehab approach in my opinion. During the acute state of damages, we've got to let the brain subside and rest. Rear crash protection is rated based on a twostep procedure. Starting points for that ratings are measurements of head restraint geometry the height of the restraint and its horizontal distance behind the rear of the head of an average size man.
If you are Patricia Tomasi and don't want your image to be displayed on 123people, please delete the image from the original source. Using our proprietary search algorithm, you can find comprehensive and centralized person related information composed of public records, phone numbers, addresses, images, Cheap Louis Vuitton Wallets videos and email addresses.
Great, we're going to freeze spending, but fully fund IDEA. What goes on to other programs in the Domestic Discretionary Plan for that to happen? Pell Grants, the LowIncome Home Energy Assistance Program, WIC, cleanenergy research, Section 8 housing dozens and a large number of programs will need to be cut by typically 6.4%.
There is a boat launching ramp for registered campers and kayaks and rowboats for rental through the park office. Look into the Nature Center for interpretive Ugg Boots Retailers programs, games, or entertainment. I've were built with a negative comment about my JapaneseAmerican truck. It came from a guy I was on a date with.