-
作者：Salemi, Peter L.; Song, Eunhye; Nelson, Barry L.; Staum, Jeremy
作者单位：MITRE Corporation; Pennsylvania Commonwealth System of Higher Education (PCSHE); Pennsylvania State University; Pennsylvania State University - University Park; Northwestern University
摘要：We consider optimizing the expected value of some performance measure of a dynamic stochastic simulation with a statistical guarantee for optimality when the decision variables are discrete, in particular, integer-ordered; the number of feasible solutions is large; and the model execution is too slow to simulate even a substantial fraction of them. Our goal is to create algorithms that stop searching when they can provide inference about the remaining optimality gap similar to the correct-sele...
-
作者：Whitt, Ward; Zhang, Xiaopei
作者单位：Columbia University
摘要：Motivated by our recent study of patient flow data from an Israeli emergency department (ED), we establish a sample path periodic Little's law (PLL), which extends the sample path Little's law (LL). The ED data analysis led us to propose a periodic stochastic process to represent the aggregate ED occupancy level, with the length of a periodic cycle being 1 week. Because we conducted the ED data analysis over successive hours, we construct our PLL in discrete time. The PLL helps explain the rem...
-
作者：Gupta, Varun; Walton, Neil
作者单位：University of Chicago; University of Manchester
摘要：We analyze join-the-shortest-queue (JSQ) in a contemporary scaling regime known as the nondegenerate slowdown (NDS) regime. Join-the-shortest-queue is a classical load-balancing policy for queueing systems with multiple parallel servers. Parallel server queueing systems are regularly analyzed and dimensioned by diffusion approximations achieved in the Halfin-Whitt scaling regime. However, when jobs must be dispatched to a server upon arrival, we advocate the nondegenerate slowdown regime to co...
-
作者：Ban, Gah-Yi; Rudin, Cynthia
作者单位：University of London; London Business School; Duke University
摘要：We investigate the data-driven newsvendor problem when one has n observations of p features related to the demand as well as historical demand data. Rather than a two-step process of first estimating a demand distribution then optimizing for the optimal order quantity, we propose solving the big data newsvendor problem via singlestep machine-learning algorithms. Specifically, we propose algorithms based on the empirical risk minimization (ERM) principle, with and without regularization, and an...
-
作者：Reich, Gregor
作者单位：University of Zurich
摘要：This paper develops a method to efficiently estimate hidden Markov models with continuous latent variables using maximum likelihood estimation. To evaluate the (marginal) likelihood function, I decompose the integral over the unobserved state variables into a series of lower dimensional integrals, and recursively approximate them using numerical quadrature and interpolation. I show that this procedure has very favorable numerical properties: First, the computational complexity grows linearly i...
-
作者：St John, Rachel; Toth, Sandor F.; Zabinsky, Zelda B.
作者单位：University of Washington; University of Washington Seattle; University of Washington; University of Washington Seattle
摘要：Wildlife corridors are often used to connect critical habitat for species protection. Mixed integer programming models have been used in the past to create wildlife corridors, but they lack the capacity to control corridor geometry. We propose an approach that employs path planning techniques from artificial intelligence to account for and control corridor geometry, such as width and length. By combining path planning with network optimization, our approach allows the user to control and optim...
-
作者：Zhou, Zhengyuan; Bambos, Nicholas; Glynn, Peter
作者单位：Stanford University; Stanford University
摘要：Power control over wireless networks has been an active area of research with significant applied impact. A well-motivated line of this research, which has received increasing attention, is applying game-theoretic tools for both gaining insight and design of algorithms. In this paper, we build on the existing work and present a simple game-theoretic formulation of power control on wireless networks that incorporates two novel features. First, we do not impose exogenous power bounds on the feas...
-
作者：Bhargava, Hemant K.; Gangwar, Manish
作者单位：University of California System; University of California Davis; Indian School of Business (ISB)
摘要：Two- (2PTs) and three-part tariffs (3PTs) are widely used for selling goods, to compensate workers, and in procurement contracts. They are practical alternatives to complex nonlinear tariffs in on-demand services and technology industries and are more profitable than the restrictive per-unit and unlimited-use pricing. A 2PT imposes both a fixed (access) fee and a per-unit (usage) fee, and a 3PT generalizes it by bundling some free units (an allowance) into the fixed fee. Intuitively, bundling ...
-
作者：Li, Jonathan Yu-Meng
作者单位：University of Ottawa
摘要：Worst-case risk measures provide a means of calculating the largest value of risk when only partial information of the underlying distribution is available. For popular risk measures such as value-at-risk (VaR) and conditional value-at-risk (CVaR) it is now known that their worst-case counterparts can be evaluated in closed form when only the first two moments are known. We show in this paper that closed-form solutions exist for a general class of law invariant coherent risk measures, which co...
-
作者：Capponi, Agostino; Cheng, W. Allen
作者单位：Columbia University
摘要：We model the decision problem faced by a profit-maximizing clearinghouse, which sets fee and margin requirements for heterogeneous traders who may default. We capture the main trade-offs underpinning the clearinghouse's choices: higher fee and better default protection come at the cost of decreased market volume. We show that the equilibrium margin requirements are determined not only by price volatility but also by trader fundamentals and funding costs. Our results (i) explain why margins are...
