2016 INFORMS Annual Meeting Program

SD03

INFORMS Nashville – 2016

3 - Discriminating Parkinson’s Disease (pd) Using Functional Connectivity And Brain Network Analysis Shouyi Wang, University of Texas at Arlington, shouyiw@uta.edu In this study, we explored the use of functional connectivity patterns in fMRI data to classify subjects on the basis of Parkinson’s disease. We explore various brain networks and features. We partition our fMRI data in 5 filtered frequency ranges. We use a proximal support vector machine paired with a minimum-redundancy and maximum-relevance feature selection method on each frequency range. We use a majority voting ensemble classification method on the results of the proximal support vector machine classification results. Our results indicate that the ensemble method is effective compared to a single broad frequency range, and that Bonferroni correction may enhance classification results. Doing Good with Good OR II Invited: Doing Good with Good OR Invited Session Chair: Chase Rainwater, University of Arkansas, 4207 Bell Engineering Center, Fayetteville, AR, 72701, United States, cer@uark.edu 1- Optimizing Breast Cancer Diagnostic Decisions to Reduce Overdiagnosis Sait Tunc, Department of Industrial and Systems Engineering, Although the early diagnosis of breast cancer through screening mammography saves thousands of lives every year, overdiagnosis of breast cancer may cause harm without benefit. We propose a comprehensive large-scale optimization model to address overdiagnosis by making better-informed diagnostic decisions and provide the exact optimal policies that potentially save up to 9% of the biopsied population from overdiagnosis.. 2 - A Decision Support System for the Management of Aid-In-Kind Donations for Turkish Red Crescent Semih Boz, Department of Industrial Engineering, Bilkent University, Ankara, Turkey, Semih Kaldirim, Bilge Kaycioglu, Buse Eylul Oruc, Eren Ozbay, Mirel Yavuz, Sinan Derindere, Ali Erkan, Pinar Ozkurt Turkish Red Crescent is the main body for collection and distribution of donations in Turkey, and this project aims to improve their donation collection and distribution processes by proposing a decision support system with numerous subsystems; e.g. implementing new process flows and their accompanying decisions. The proposed system is approved by TRC and it is being integrated to their systems. SD04 101D-MCC Optimizing Urban Infrastructure Resilience Under Climate Change Sponsored: Energy, Natural Res & the Environment I Environment & Sustainability Sponsored Session Chair: Mohammad Ramshani, University of Tennessee, 524 John D. Tickle Building, Knoxville, TN, 37996, United States, mramshan@vols.utk.edu 1 - Optimizing Green Roof Integrated Photovoltaics Placement Under Climate Change Mohammad Ramshani, University of Tennessee, mramshan@vols.utk.edu, Xueping Li, Anahita Khojandi, Olufemi A Omitaomu We develop a two-stage stochastic model to optimally place green roofs and/or Photovoltaic panels under climate change uncertainty, with the aim of improving urban system resilience. Different climate forecasts from different climate models are taken into account as different scenarios. The interaction between green roofs and Photovoltaic panels in terms of efficiency is considered. An efficient L-shaped algorithm is developed. Computational studies along with sensitivity analysis are conducted. University of Wisconsin-Madison, Madison, WI 53706, stunc@wisc.edu, Oguzhan Alagoz, Elizabeth S. Burnside SD03 101C-MCC

2 - Optimal Planning Of Green Infrastructure Placement Under Precipitation Uncertainty Masoud Barah, University of Tennessee, mbarah@vols.utk.edu, Anahita Khojandi, Xueping Li, Jon Hathaway Green Infrastructures (GIs) are low cost, low regret strategies that can dramatically contribute to stormwater management. We develop a multi-objective stochastic programming model to determine the optimal placement of GIs across a set of candidate locations in a watershed to minimize the excess runoff under short-term and medium-term precipitation uncertainties. We calibrate the model using precipitation projections and stormwater system’s hydrologic responses to them. We obtain the optimal GI placement for a watershed and perform sensitivity and robustness analyses to provide insights. 3 - Optimal Placement Of Green Infrastructure Under Uncertainty Anahita Khojandi, University of Tennessee, 603 W Main St., Apt 801, Knoxville, TN, 37902, United States, anahitakhojandi@gmail.com, Mohit Shukla, Xueping Li, Mohammad Ramshani Despite the environmental and societal benefits of Green Infrastructure (GI), they are mostly planned and established in response to an existing problem rather than being actively incorporated into the early stages of urban planning. In this paper, we present a stochastic model that would allow urban planners to incorporate uncertainties in population and climate predictions, land use and budgetary constraints and the ‘connectivity’ between GIs into the decision making process of GI placement on a county or city scale land area. The proposed approach is tested on data from a real county to evaluate its utility. SD05 101E-MCC ENRE Award Session Sponsored: Energy, Natural Res & the Environment, Energy I Electricity Sponsored Session Chair: Andy Sun, Georgia Institute of Technology, Atlanta, GA, 30332, United States, andy.sun@isye.gatech.edu SD06 102A-MCC INFORMS 2016 Data Mining Best Student Paper Awards II Sponsored: Data Mining Sponsored Session Chair: Mustafa Gokce Baydogan, Bogazici University, Istanbul, Turkey, baydoganmustafa@gmail.com SD07 102B-MCC Joint Session DM/AI: Predictive Analytics in Data Science Sponsored: Data Mining Sponsored Session Chair: Xi Wang, University of Iowa, S210 John Pappajohn Business Building, Iowa City, IA, 52242-1000, xi-wang-1@uiowa.edu 1 - Link Prediction In Multi-relational Networks For Online Health Communities Xi Wang, The University of Iowa, xi-wang-1@uiowa.edu Kang Zhao Online Health Communities (OHCs) are a popular resource for those with health problems to exchange information and support. Users often interact via multiple communication channels, such as online discussions, blogs, and private messages. Connections among users via different channels form a multi-relational social network. Using data from a smoking cessation network, this study aims to predict links between users in one sub-network based on information from other sub- networks. Our findings regarding tie formation will inform the development and ongoing management of online health communities.

96