Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MB55

n MB55 North Bldg 232C Advancing Diversity of Women in STEM Sponsored: Women in OR/MS (WORMS) Sponsored Session Chair: Sudharshana A. Apte, Altria Client Services, Richmond, VA, 23221, United States 1 - Four Things Women in Data Science Can Learn from Game of Thrones Jennifer Priestley, Kennesaw State University, Kennesaw, GA, United States Studies consistently find that women are underrepresented in most computational disciplines - particularly in Analytics and Data Science. And although events and organizations that encourage girls K-12 to learn to code have increased over the last few years, the number of college-age women in computational disciplines has not increased. Nor has the proportion of women in analytical leadership positions. In this talk, one of the few female directors of a Ph.D. in Data Science will provide perspective on how the discipline can attract (and retain) more female talent. These points will be framed through the popular HBO Series Game of Thrones. 2 - Unintended Consequences of Increasing Female Engineers’ Representation in Managerial Roles Teresa Cadrador, University of Illinois at Urbana Champaign, Champaign, IL, United States Engineering remains one of the most highly and persistently sex segregated occupations in the US. Though the extant literature submits that women’s increased access to managerial positions in male-dominated occupations should represent an important strategy for addressing sex segregation, my recent research suggests that women’s representation in managerial roles in engineering may promote the very sex segregation it is attempting to mitigate. I will present the results of two field studies which highlight how and why female engineers’ movement into managerial roles fosters a form of intra-occupational sex segregation with unintended (and largely negative) consequences for women. 3 - Advancing Diversity: Ideas for Faculty Searches Karen Smilowitz, Northwestern University, Industrial Engineering Management Scienc, 2145 Sheridan Road RM D239, Evanston, IL, 60208, United States In this talk, I will present related research, tips and lessons learned to advance diversity in STEM through more structured faculty searches. n MB56 West Bldg 101A HAS Distinguished Scholar Lecture Sanjay Mehrotra Sponsored: Health Applications Sponsored Session Chair: Ebru Korular Bish, Virginia Tech, Blacksburg, VA, 24060, United States Co-Chair: Tinglong Dai, Johns Hopkins University, Johns Hopkins University, Baltimore, MD, 21202, United States 1 - Big Data Analytics and Operations Research Modeling: Broader Implications of Lessons Learned from a Journey towards Influencing Policy Change in Organ Transplant Sanjay Mehrotra, Northwestern University, Dept of I. E. / M. S. C246 Tech Inst, 2145 Sheridan Road, Evanston, IL, 60208-3119, United States As health applications scientists we excel in model development and establishing their properties. We will reverse this order. We will start with data and its analysis to establish the problem. We will next search for natural evidence to direct model development. We will need to establish our credibility among stake holders. Only now we can develop optimization models that leverage OR insights. But, they have to outperform an existing proposal through dominance on multiple metrics. It is now time to give up optimality, to help increase solution acceptability. Insights will lead us to developing novel multi-objective ambiguous optimization methodologies. This will be done using organ transplant as a case study. The newly developed methodologies and modeling paradigms have wide applications.

n MB57 West Bldg 101B Disease Modeling and Decision-support Tools for Medical Decision Making Sponsored: Health Applications Sponsored Session Chair: Ethan Mark, Georgia Tech, Atlanta, GA, 30363, United States Co-Chair: Pinar Keskinocak, Georgia Institute of Technology, Atlanta, GA, 30332, United States 1 - An Empirical Analysis of the Opioid Prescription Epidemic Alireza Boloori, Arizona State University, Tempe, AZ, 85283, United States, Soroush Saghafian, Stephen Traub Opioid epidemic has been largely attributed to the overprescription of opioid painkillers. As a result, many medical guidelines have recently urged healthcare providers to lessen opioid prescriptions in their medical practices. This, however, could negatively affect those patients who suffer from acute or chronic pain symptoms. Utilizing commercial insurance claims and encounters data, we first analyze the trade-off between the opioid epidemic and pain management. Based on our results, we then provide recommendations for both policymakers and healthcare providers. 2 - Predictive Modeling of Multiple Chronic Conditions Development Adel Alaeddini, University of Texas at San Antonio, Department of Mechanical Engineering, One UTSA Circle, San Antonio, TX, 78249, United States Development of multiple chronic conditions follows a complex process, influenced by several factors including the inter-relationship of the existing conditions, patient-level risk factors, etc. Using a large retrospective dataset of patient population, we build a machine learning model to explore the patterns of multiple chronic conditions development. 3 - Computer-aided Diagnostic Models for Breast Cancer Screening and Diagnosis Mammography is used for early breast cancer screening & diagnosis whereas mammography interpretation a difficult task due to similarities between early signs of breast cancer and normal structures in images. We have developed computer-aided diagnostic models and tested them using real-life data from private mammography databases. In this presentation, we describe our experiences, list the limitations of the existing CADx models, and provide possible future research directions. 4 - Leveraging TCGA Gene Expression Data to Build Predictive Models for Cancer Drug Response Toyya Pujol, Georgia Institute of Technology, Atlanta, GA, 30318, United States, Evan Clayton, Peng Qiu Personalized oncology promises to increase the success rate of cancer drug therapy by using molecular tumor profiles to determine the optimal therapeutic for an individual. Here, we build machine learning models using gene expression data directly from patients’ primary tumor tissues to predict whether a patient will respond positively or negatively to a given drug. We show our models predict, with up to 86% accuracy, whether a patient will be a responder or not, and discuss how our findings may aid oncologists with making critical treatment decisions. 5 - Using Machine Learning and Simulation to Compare Increased Organ Transplant Survival to Waiting for a Non-Increased Risk Organ Ethan Mark, Georgia Tech, 213 16th Street, NW, Apartment 4, Atlanta, GA, 30363, United States Abstract not available. Oguzhan Alagoz, University of Wisconsin-Madison, 3242 Mechanical Engineering Building, 1513 University Avenue, Madison, WI, 53706, United States

168