2015 Informs Annual Meeting

WB78

INFORMS Philadelphia – 2015

WB78 78-Room 301, CC Natural Resources Contributed Session

WB79 79-Room 302, CC

Software Demonstration Cluster: Software Demonstrations Invited Session 1 - FICO – Turnkey Optimization on the Cloud Oliver Bastert, Ph.D., FICO, oliver.bastert@fico.com

Chair: Wanshan Zhu, Associate Professor, Tsinghua University, Shunde Building 613,, Depart of Industrial Engineering, Beijing, 100084, China, zhuws@tsinighua.edu.cn 1 - A Model to Assess the Impact of Overemployment and Subsidized Fuel Price on National Oil Companies Sergio Cabrales, Visiting Professor, Universidad de los Andes, Carrera 53A No 127 -30 apt:1204, Bogota, DC, 0111111, Colombia, s-cabral@uniandes.edu.co, Juan Banavides, Rafael Bautista National Oil Companies (NOCs) produce 61% of worldwide oil. International Oil Companies (IOCs) maximize the expected net present value of their profits, whereas for NOCs are not necessarily the only objective. Indeed, NOCs’ objectives often include non-commercial goals such as employment and fuel subsidies. We develop an optimal control model to estimate the impact of overemployment and fuel subsidies in terms of market value, production, and reinvestment. 2 - Strategic Bidding for a Virtual Power Plant: A Price-taker Robust Optimization Approach Luis Baringo, Assistant Professor, University of Castilla-La Mancha, E.T.S.I.Industriales, Avda. Camilo Jose Cela s/n, Ciudad Real, 13071, Spain, Luis.Baringo@uclm.es, Morteza Rahimiyan We consider an energy management system that controls a cluster of price- responsive demands, a wind-power plant and an energy storage facility that are interconnected within a small size electric energy system and that constitute a virtual power plant (VPP). We propose a two-stage robust optimization approach for the strategic bidding of this VPP in the day-ahead and the real-time markets. Uncertainties in wind-power production and market prices are represented through confidence bounds. 3 - Faustmann Revisited and Updated with Modern Operations Research Tools Gysbert Wessels, Consultant, 3748 Bay Tree Pl, Blacksburg, VA, United States of America, wessels.gys@gmail.com Faustmann’s article published in 1849 remains an important contribution to Forestry Economics and Management. What if Faustmann had access to the current tools of OR/MS? The presentation will show how “Faustmann’s Formulae” can be updated and generalized. It will be shown that Faustmann’s calculations and the generalized approach give the same results under Faustmann’s assumptions. Many of the assumptions can be relaxed, making the generalized approach practical to use in modern forestry. 4 - Stackelberg vs Cournot in a Natural Resource Oligopoly Luca Colombo, Deakin University, Burwood Campus, In this paper, we compare and contrast feedback Nash and Stackelberg equilibrium strategies in a differential oligopoly game in which production requires exploitation of a common-property renewable resource. We find that the the incentives for firms to take the lead are higher the higher the stock of the resource. We also find that the Cournot equilibrium can be more efficient than the Stackelberg equilibrium, both in the short-run and at the stationary equilibrium. 5 - Electricity Capacity Planning with Cross Border Exchange Wanshan Zhu, Associate Professor, Tsinghua University, Shunde Building 613,, Depart of Industrial Engineering, Beijing, 100084, China, zhuws@tsinighua.edu.cn This paper studies the impact of cross border maximum exchange capacity on the optimal electricity generation capacity mix. We show that models with exogenous spot prices miss the influence of capacity expansion decisions on the market, and suggests a model where spot prices are endogenous, as a function of demand and available capacity. A case study in France with the EPEX Spot prices is made. 221 Burwood Highway, Melbourne, Australia, luca.colombo@deakin.edu.au, Paola Labrecciosa

In this workshop, we will demonstrate enhancements for modeling and solving linear, mixed integer and nonlinear optimization problems using the latest release of FICO® Xpress version 7.9. We will show how to rapidly turn optimiza- tion models into collaborative applications deployed on the FICO® Analytic Cloud. Also, learn how these capabilities can be combined with analytic model- ing and decision rules to deliver powerful cloud-based or on-premises decision management solutions via the FICO® Decision Management Suite. 2 - IBM Academic Initiative Group - IBM Academic Initiative for Cloud...Building Next-generation Skills IBM Academic Initiative Group Digital transformations are requiring students, no matter what their major, understand how to leverage and build solutions on the cloud. With the new Academic Initiative for Cloud offer, faculty and students can have hands-on cloud-based experiences to propel radical ideas and innovation using IBM Watson, Internet of Things, big data, analytics, mobile and more. Join us to col- laborate on best practices to empower innovation in the classroom and beyond! http://ibm.biz/aiforcloud. Wednesday, 12:45pm - 2:15pm WC01 01-Room 301, Marriott Military Cognitive Analysis, Value based Acquisition and Military Recruiting Prediction Models Sponsor: Military Applications Sponsored Session Chair: Mike Teter, Ltc, US Army, 515 Michelson Rd, Monterey, CA, 93940, United States of America, Michael.d.teter6.mil@mail.mil 1 - Predicting Market Depth for Military Recruiting Jon Alt, Assistant Professor, Naval Postgraduate School, Department of Operations Research, Naval Postgraduate School, Monterey, CA, 93943, United States of America, jkalt@nps.edu, Sam Buttrey This ongoing research demonstrates the application of statistics and machine learning to identify those geographic areas that are more likely to produce military recruits. It also seeks to compare factors influencing service specific market depth. This comparison may inform the development of a common decision support framework. Practical difficulties in preparing and open source data for this purpose are discussed. 2 - Utilizing Socio-economic Factors to Evaluate Recruiting Potential for a US Army Recruiting Company US Army currently calculates recruiting capacity as a four year weighted average of historical data. We investigate two alternate methods for the same task — multiple linear regression (MLR) and Poisson regression (PR). Regression methods can account for the impact of economic factors on recruiting capacity, whereas weighted average methods do not. Surprisingly, we show that MLR models provide better fits than PR models, even though existing literature largely focuses on PR models. 3 - Military Modification of The Iowa Gambling Task and Wisconsin Card Sorting Task Cardy Moten, Maj, TRADOC Analysis Center-Monterey, 700 Dyer Sandra Jackson, US Army, Thayer Hall, West Point, United States of America, jackson.sandra.y@gmail.com, Nedialko Dimitrov, Jon Alt

Road, Room 183, Monterey, CA, 93943, United States of America, cmoten@nps.edu, Quinn Kennedy, Jon Alt, Peter Nesbitt

TRAC-Monterey and the Naval Postgraduate School (NPS) have developed tasks to measure military decision-making performance modeled after the Iowa Gambling and Wisconsin Card Sorting Tasks. These tasks focus on high stakes and uncertain environments particular to military decision making conditions. Thirty- four officers were tested on their levels of cognitive flexibility and reinforcement learning. This presentation will discuss task development, validation, and insights of measured components.

426