Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

SD32

3 - Trip Purpose Inference and Prediction with Social Media Data and Google Places Yu Cui, University at Buffalo, The State University of New York, 314 Bell Hall, Buffalo, NY, 14260, United States, Chuishi Meng, Qing He, Jing Gao This research has addressed the problem of trip purpose prediction with both Google Places and social media data. First, this paper provides a new approach to match Point of Interests (POIs) from Google Places API with historical Twitter data. Therefore, the popularity of each POI can be obtained. Moreover, a Bayesian neural network (BNN) is employed to model the trip dependence within each individual’s daily trip chain and infer the trip purpose. 4 - Quantifying Disruption Impact Across Transportation Networks Priyadarshan Patil, The University of Texas at Austin, Austin, TX, United States, Steve Boyles, William Alexander This research quantifies the propagation of impact on a transportation network, caused by a disruption on the network. We propose to model the impact using explanatory variables such as network characteristics, flow patterns, and link interactions. A classical statistical regression-based approach is contrasted with a method based on neural networks, applied and evaluated on real-world urban networks along with megaregional networks. The results assist in resource allocation and rapid response to network disruptions. n SD32 North Bldg 222B Joint Session TSL/MIF: Sustainable City Logistics – Improving Efficiency of Movement of Freight Sponsored: TSL/Freight Transportation & Logistics Sponsored Session Chair: Diana Gineth Ramirez-Rios, Rensselaer Polytechnic Institute, Troy, NY, 12180, United States 1 - Assessing Impacts of Emerging Market and Technological Trends for Freight-efficient Land-uses Sofia Perez-Guzman, Rensselaer Polytechnic Institute, Troy, NY, 12180, United States, Jose Holguin-Veras, Diana Gineth Ramirez- Rios, Juvena Ng, Wilfredo Yushimito Del Valle, Joshua Schmid This study proposes an assessment of market trends and emerging technologies, and how the impacts brought by these disruptors affect the planning of freight- efficient land uses. To this extent, an evaluation framework was developed, which incorporates how trends affect decisions of stakeholders involved in different supply chains. The framework systematically captures each one of the decisions made by stakeholders based on the minimization of their total social costs (i.e. private costs and externalities) and translates these into short and long-term effects and land-use measures. 2 - A Spatial Methodology for Characterizing Freight in Metropolitan Areas on the US Carlos Rivera-Gonzalez, Rensselaer Polytechnic Institute, Troy, NY, United States, Jose Holguin-Veras, Juvena Ng, Diana Gineth Ramirez-Rios This study proposes an innovative spatial methodology for understanding the economic activities in Metropolitan Statistical Areas (MSAs). Through the development of a spatial index and the incorporation of freight activity estimated at a ZIP code level, the study was able to identify freight activity clusters and freight corridors that could be useful for the decisions on the allocation of facilities in different urban settings. The analysis responds to a selected group of MSAs in the United States that represent a diversity of urban areas in terms of size and geographic location. 3 - State of the Art and Practice of Urban Freight Management Johanna Amaya Leal, Iowa State University, 2167 Union Drive, Room 3127, Ames, IA, 50011, United States, Jose Holguin-Veras, Ivan Sanchez-Diaz, Michael Browne, Jeffrey Wojtowicz This paper conducts a review of the public-sector initiatives that could be used to improve freight activity in metropolitan areas; collects data about initiatives that have been implemented and their performance; and produces a ranking of suggested initiatives. The characterization and performance of the initiatives was based on a survey that collected data from countries and cities throughout the world. The paper ends with a discussion of chief findings. 4 - Patrol Police Routing It is well-known that police patrolling is one of the best preventive practices for public safety against urban crimes. This work, deals with the problem of planning police patrol routes to minimize the overall risk at minimum cost. A specific mathematical formulation, models the problem under critical time constraints and resources. Algorithms of ant colony and evolutionary techniques, offers effective solutions for this model. A case study in Barranquilla (Colombia), allows validate the performance of our approach in real scenarios. Ruben Yie, Universidad el Norte, Km 5 Antigua via Puerto Colombia, Barranquilla, Colombia, Andrea Margarita Ditta

n SD33 North Bldg 222C George B Dantzig Dissertation Prize I Emerging Topic: George B Dantzig Dissertation Prize Emerging Topic Session Chair: Mor Armony, New York University, 44 West 4th Street #8-62, New York, NY, 10012, United States 1 - Models and Algorithms for Transportation in the Sharing Economy Daniel Freund, Cornell University, Ithaca, NY, USA. This thesis consist of two parts. The first deals with bike-sharing systems which are now ubiquitous across the U.S.A. We have worked with Motivate, the operatorofthesystemsin,forexample,NewYorkCity,Chicago,andSanFrancisco,to innovate a data-driven approach to managing both their day-to-day operations and to provide insight on several central issues in the design of their systems. This work required the development of a number of new optimization models, characterizing their mathematical structure, and using this insight in designing algorithmstosolvethem. Manyoftheseprojectshavebeenfullyimplementedto improve the design, rebalancing, and maintenance of Motivate’s systems across the country. 2 - Optimization and Machine Learning Methods for Diagnostic Testing of Prostate Cancer Selin Merdan, PhD Candidate, University of Michigan, Ann Arbor, MI, USA. Technological advances in biomarkers and imaging tests are creating new avenues to advance precision health for early detection of cancer. These advances have resulted in multiple layers of information that can be used to make clinical decisions, but how to best use these multiple sources of information is a challenging engineering problem due to the high cost and imperfect sensitivity and specificity of these tests. Questions that need to be addressed include which diagnostic tests to choose and how to best integrate them, in order to optimally balance the competing goals of early disease detection and minimal cost and harm from unnecessary testing. To study these research questions, we present new optimization-based models and data-driven analytic methods in three parts to improve early detection of prostate cancer (PCa). 3 - Latent Variable Model Estimation via Collaborative Filtering Christina Lee Yu, Cornell University, Ithaca, NY, USA. Similarity based collaborative filtering for matrix completion is a popular heuristic that has been used widely across industry in the previous decades to build recommendation systems, due to its simplicity and scalability. However, despite its popularity, there has been little theoretical foundation explaining its widespread success. In this thesis, we prove theoretical guarantees for collaborative filtering under a nonparametric latent variable model, which arises from the natural property of “exchangeability”, i.e. invariance under relabeling of the dataset. The analysis suggests that similarity based collaborative filtering can be viewed as kernel regression for latent variable models, where the features are not directly observed and the kernel must be estimated from the data. In addition, while classical collaborative filtering typically requires a dense dataset, this thesis proposes a new collaborative filtering algorithm which compares larger radius neighborhoods of data to compute similarities, and show that the estimate converges even for very sparse datasets, which has implications towards sparse graphon estimation.

n SD34 North Bldg 223 4:30 - 5:15 Elsevier/5:15 - 6:00 Wayfair Vendor Demo Session 1- How to Get Published and Meet the Editors Session Simon Jones, Elsevier, Oxford, United Kingdom

How to get Published in Operations Research Journals addresses how to prepare and submit a manuscript using correct manuscript language, and how to structure an article. 2- Geo-Testing for Real-World Marketing Nathan Vierling-Claassen, Wayfair, Boston, MA, United States Optimized geographic testing is becoming a popular method for measuring the value of marketing interventions that are not amenable to more standard controlled testing methods, but getting the method to work in a real-world business environment is not always straightforward. This tutorial, we will give an overview of geo-testing methods at Wayfair. Using e-commerce marketing as a case study, we will cover practical approaches including: When & why to choose geo-testing as opposed to other more standard techniques. Test design & optimization to generate national results from local tests Quantifying & avoiding “geo-leakageö. Test design methods for cost reduction while retaining statistical power.

102