Informs Annual Meeting 2017

TB57

INFORMS Houston – 2017

TB57

fuel type or implement an emission control device to comply with the sulphur emission limits. Each of these choices will eventually have advantages and disadvantages in terms of cost and efficiency. This paper evaluates the selection of relevant control strategy for various kinds of ships. 2 - Berth Allocation – Do we have it all Wrong? Stefan Voss, University of Hamburg, Von-Melle-Park 5, Hamburg, 20146, Germany, stefan.voss@uni-hamburg.de The Berth Allocation Problem (BAP) aims at assigning and scheduling incoming vessels to berthing positions along the quay of a container terminal. During the last decades many terminal operators and academics have developed algorithms and decision support systems for optimizing berth planning including the BAP. Despite a wealth of existing systems as well as academic surveys there are still quite a few issues that seem to be neglected including the consideration of berth error as well as disturbances in the daily operation. Here, we set up a research agenda pointing out that we need to consider and model disturbances and stochastic influences to make the problem more realistic. 3 - Geometric Partitioning Algorithms for Facility Location Problems Mehdi Behroozi, Assistant Professor, Northeastern University, Boston, MA, 02115, United States, m.behroozi@neu.edu, Reyhaneh Mohammadi We present fast approximation algorithms for the continuous facility location problems such as k-centers problem and k-medians problem over a convex polygon. The objective here is to find the location of k facilities so as to minimize a distance-based cost function. Our strategy is to use a geometric partitioning scheme to partition the convex polygon into k sub-regions and allocate one facility to each sub-region. 4 - Synchronized Scheduling of Yard Cranes and Vehicles in Container Yards Chenhao Zhou, National University of Singapore, 1 Engineering Drive 2, E1A-05-21, Singapore, 117576, Singapore, a0096340@u.nus.edu, Haobin Li, Byung Kwon Lee, Loo Hay Lee, Ek Peng Chew In the container block, the handshake between a yard crane and a vehicle when handling over a container can lead to unproductive waiting time, and consequently low block throughput. This study aims to discuss the synchronized scheduling to reduce the waiting time and to improve the handling efficiency between yard cranes and vehicles. The decision elements are the handshake positions for handshaking, and the operations sequence of activities, under the Chebyshev metric of yard crane movement. The scheduling algorithm is proposed and examined its effectiveness on a simulation testbed. 5 - Modeling MIAPP NALT in Order Picking Applications with a Class Based Storage Policy Jingming Liu, Research Assistant, University of Arkansas, 920 N Leverett, Apt 811, Fayetteville, AR, 72701, United States, liujingming1215@gmail.com, Haitao Liao, John A. White A multiple-in-the-aisle pick positions (MIAPP) case picking operation is modeled as an M|Ek|1 queueing problem. In the queueing formulation, the narrow aisle lift truck (NALT) serves two types of customers: single-command storage, single- command retrieval. The travel time probability distributions are derived under a rectilinear travel and class-based storage policy, and an Erlang distribution is used to approximate the distribution of service times. 362F Joint session RAS/Practice: Intermodal Transportation Sponsored: Railway Applications Sponsored Session Chair: Nathaniel Richmond, University of Iowa, The Colony, TX, 75056, United States, nathaniel-richmond@uiowa.edu Co-Chair: Mike D. Prince, BNSF Railway, BNSF Railway, Fort Worth, TX, 76244, United States, mike.prince@bnsf.com 1 - Breaking Deadlocks in Widespan Crane Planning Steven Tyber, Operations Researcher, GE Transportation, 500 W. Monroe St, Chicago, IL, 60613, United States, tyber@ge.com We explore the problem deadlocks in planning wide span crane operations. Like many scheduling problems, wide span crane tasks have precedence relations that need to be observed to produce a feasible plan. When a set of tasks are mutually dependent, they are said to be deadlocked and prevent any plan from being feasible. In this talk, we show how such deadlocks arise in wide span crane operations and how standard practices for arranging and loading units give rise to an easily executed solution to this challenging problem. TB59

362D Statistical and Optimal Learning Sponsored: Simulation Sponsored Session

Chair: Ilya O. Ryzhov, University of Maryland, 4322 Van Munching Hall, University of Maryland, College Park, MD, 20742, United States, iryzhov@rhsmith.umd.edu 1 - Optimization of Tree Ensembles Velibor Misic, University of California Los Angeles, Los Angeles, CA, United States, velibor.misic@anderson.ucla.edu Decision trees and tree ensemble models such as random forests are often used to predict the effect of different decisions. While such models are widely used for predictive purposes, little is known about how to use them for effective decisions. In this talk, we consider the problem of finding a decision that is optimal with respect to a tree ensemble model. We formulate this problem as a mixed-integer optimization model and present theoretical results on the structure of this model. We demonstrate the value of our methodology on real datasets and present two detailed applications to drug design and multiproduct pricing. 2 - Markov Decision Processes with Exogenous Variables Robert Louis Bray, 830 Hinman Ave., 2s, Evanston, IL, 60202, United States, robertlbray@gmail.com I present two new dynamic program solution methods: endogenous value iteration and endogenous policy iteration. I define the exogenous space of a Markov decision process as the vector space comprising the set of value function deviations that never influence the policy function. My algorithms iteratively apply Bellman contractions, projecting the portion of the value function that lies in the exogenous space to zero at each step. When the exogenous space is one- dimensional, my algorithms are equivalent to relative value iteration and relative policy iteration; but when the exogenous space is multi-dimensional, my algorithms can be faster. 3 - Derivative-based and Derivative-free Adaptive Sampling Trust Region Optimization Raghu Pasupathy, Purdue University, West Lafayette, IN, United States, pasupath@purdue.edu, Susan R.Hunter, Sara Shashaani, Fatemeh S. Hashemi We will present an adaptive sampling trust region method for solving unconstrained stochastic optimization problems where the objective function can only be estimated using an inexact oracle, e.g., Monte Carlo or quasi-Monte Carlo simulation. We will discuss both derivative-based and derivative-free versions of the algorithm, along with convergence calculations. We will also briefly demonstrate a global complexity result in a specific deterministic inexact oracle setting. 4 - Rate-optimality of the Complete Expected Improvement Criterion Ye Chen, University of Maryland, 1305 Mathematics Building, University of Maryland, College Park, MD, United States, yechen@math.umd.edu, Ilya O.Ryzhov Expected improvement (EI) is a leading algorithmic approach to simulation-based optimization. However,it was recently proved that, in the context of ranking and selection, some of the most well-known EI-type methods cause the probability of incorrect selection to converge at suboptimal rates. We investigate a more recent variant of EI (known as “complete EI”) that was proposed by Salemi, Nelson, and Staum (2014), and summarize results showing that, with some minor modifications, complete EI can be made to achieve the optimal convergence rate in ranking and selection with independent Gaussian noise. This is the strongest theoretical guarantee available for any EI-type method. 362E Applications of Facility Logistics - I Sponsored: TSL, Facility Logistics Sponsored Session Chair: Amir Gharehgozli, Texas A & M University, Galveston, TX, 77553, United States, gharehga@tamug.edu 1 - Emission Control Strategies in the Maritime Transport: LUG vs. Scrubber Debate Okan Duru, Nanyang Technological University, 50 Nanyang, N1-01C-95, Singapore, 638798, Singapore, duruokan@yahoo.com This paper investigates the trade-off between three major strategies for controlling Sulphur emissions from ships due to the new international regulations. According to the recent international regulation, shipping firms will need to change their TB58

324