Informs Annual Meeting 2017

SD77

INFORMS Houston – 2017

SD77

SD78B

372F Recent Advances in Computational Optimization Sponsored: Optimization, Computational Optimization and Software Sponsored Session Chair: Robert J Vanderbei, Princeton University, Princeton, NJ, 08544, United States, rvdb@princeton.edu Co-Chair: John W Chinneck, Carleton University, Systems and Computer Engineering, 1125 Colonel By Drive, Ottawa, ON, K1S 5B6, Canada, chinneck@sce.carleton.ca 1 - Fast Approximate Solution of Very Large Linear Programs John W. Chinneck, Carleton University, Systems and Computer Engineering, 1125 Colonel By Drive, Ottawa, ON, K1S.5B6, Canada, chinneck@sce.carleton.ca It can be useful to find an approximate solution quickly for huge linear programs, e.g. for screening options or providing an advanced start for an accurate solver. I present a method that optimizes a combined objective function that has a quadratic penalty term for constraint violations, a movement direction based on simultaneous projection and a quadratic approximation to the combined objective function. Initial points are scattered over a gradually shrinking box surrounding the current best solution. Empirical results show that the method often provides good approximate solutions in seconds for models which require hours of computation for an accurate solution via commercial LP solvers. 2 - High Performance Computing with GAMS Frederik Fiand, GAMS.Software GmbH, Augustinusstr. 11b, Frechen-Koenigsdorf, Germany, ffiand@gams.com, Michael Bussieck BEAM-ME is a project addressing the need for new and improved solution approaches for energy system models of vast size. The focus is on large-scale LPs having a block structure that is not well exploited by state-of-the-art LP solvers. Without considering this structure the models quickly become intractable. We develop new solution algorithms that exploit the block structure and utilize HPC environments. We discuss extensions to GAMS that allow annotations specifying block structure, as well as other challenges overcome to produce some initial results. 3 - Geometric Visualization of Linear Programming Algorithms Robert J.Vanderbei, Princeton University, Operations Research & Financial Engineering, 209 Sherrerd Hall, Princeton, NJ, 08544, United States, rvdb@princeton.edu Algorithms for linear programming can be described both algebraically and geometrically. While the geometrical view is highly compelling, it is hard to grasp in dimensions higher than three. In this talk, I will provide a geometric perspective to some optimization algorithms as they apply to problems that have thousands of variables but the stem from a physical problem in two or three dimensional space. Viewing the progression of these algorithms in this highly comprehensible low dimensional space provides some interesting insights into the nature of the various algorithms. 381A ENRE Award Session Sponsored: Energy, Natural Res & the Environment Electricity Sponsored Session Chair: Andy Sun, Georgia Institute of Technology, Atlanta, GA, 30312, United States, andy.sun@isye.gatech.edu ENRE will announce the winners of its Best Publication Awards in Environment and Sustainability, Energy, and Natural Resources as well as the Young Researcher Prize and the Student Best Paper Award at this session. The winners will be invited to give presentations on their work. SD78

380B Analytic Hierachy Process Contributed Session Chair: Xiaojun (Gene) Shan, UHCL, Houston, TX, United States, shan@uhcl.edu 1 - Sustainable Public Administration Yuji Sato, Professor, Chukyo University, 101 Yagotohonmach, Showa, Nagoya, Aichi, 466-8666, Japan, ysatoh@1988.jukuin.keio.ac.jp This paper considers sustainable public administration. Based on the property of public services and the characteristic of public sector quantified by the AHP, the best matching between public services and their responsible organization will be considered, in which properties of each public service and characteristics of alternative sectors is quantified. The results imply the scheme of sustainable public administration. 2 - Comparing Traditional, Online and Hybrid MBA with Analytic Hierarchy Process Xiaojun (Gene) Shan, University of Houston Clear Lake, Houston, TX, United States, shan@uhcl.edu, Saranya Sarika This paper compares the different modes of instruction in the MBA class (traditional, online, and hybrid) with analytic hierarchy process. The important criteria for the comparison were identified based on literature review and then a survey was conducted to derive sub criteria considering the previous research as well as the students’ opinions. The survey data was analyzed and was applied to the criteria and sub criteria. The best mode of MBA instruction was determined to be traditional teaching. This supports that though technology has developed beyond imagination, classroom teaching is still the best medium for a student to understand, participate and learn from the MBA class. 3 - Selection of Solid Waste Management Program for the University of Houston - Clear Lake using Analytic Hierarchy Process Xiaojun (Gene) Shan, University of Houston Clear Lake, Houston, TX, United States, shan@uhcl.edu, Krishna Deep Kambhampati As a university grows, there is an increase in waste generated on campus. If the waste is not properly disposed of, it could lead to health hazards for both humans and other lives and disrupt the ecosystem of the campus. Therefore, it is important for educational institutions to implement a Solid Waste Management (SWM) program. This paper studies the different kinds of SWM programs that are currently in use. The objective of this paper is to identify the best SWM practice suitable for the University of Houston Clear Lake (UHCL). To evaluate the different SWM practices, the Analytic Hierarchy Process (AHP) is used. The results show that waste reduction is the most appropriate SWM program for UHCL. 4 - Courier Delivery Services Company Selection using Analytical Hierarchy Process and Data Envelopment Analysis Xiaojun (Gene) Shan, University of Houston Clear Lake, Houston, TX, 77058, United States, shan@uhcl.edu, Sindhuja Yerroju Business owners face the challenge of choosing the best third-party logistics provider among many alternatives. This problem could be formulated as a multi- criteria decision making problem. In this study, we apply analytical hierarchy process (AHP) to select the best provider in terms of the types of services provided, the number of depots, and delivery mode. The results from the analytical hierarchy process are evaluated against the results of data envelopment analysis (DEA). In DEA, the various Decision Making Units (DMU’s) are evaluated considering number of depots, number of employees and amount of equipment as inputs and revenue as an output. 5 - Using Analytical Hierarchy Process (AHP) to Select the Best Thermo Scientific Co2 Incubator Xiaojun (Gene) Shan, UHCL, houston, TX, United States, shan@uhcl.edu, Juilee Shukla Incubators are used in the tissue culturing laboratories to maintain a stable environment for the growth of cells, cultures and antibodies. Since incubators involve significant investment, it is important to select the best incubator for the purpose. The purpose of this study is to find the best incubator among the four Thermo Scientific CO2 Incubators, using Analytical Hierarchy Process (AHP) with five criteria. Each Incubator has its own range of precision to control the temperature suitable for tissue cell growth, humidity to maintain a stable environment, CO2 range to avoid the desiccation of the cells and the number of shelves, which signifies the storage volume available inside the incubator.

128