INFORMS 2021 Program Book

INFORMS Anaheim 2021

MD26

2 - Business Method Innovation in US Manufacturing and Trade Tian Chan, Emory University’s Goizueta Business School, GA, 30030, United States, Anandhi S. Bharadwaj, Deepa Varadarajan “Kx” “What kind of business method innovation do firms in the manufacturing and trade sectors engage in? Does engagement in business method innovation create value? Using classification and text analysis of the business method patents, we show that business method innovation in these sectors is primarily aimed at improving the ways in which tangible products are marketed, delivered, or enhanced through service offerings. Leveraging the exogenous shock of the State Street ruling, which first recognized business methods as a patentable category, we show that the value of firms with business method patents relative to comparable peers with no such patents to be higher by 9% after State Street. We further show that manufacturers gained a smaller 7% increase, relative to a 25% gain for firms in the trade sectors; and only firms with broad innovation scope see a significant value bump. 3 - The Effect of Routine Communication Within and Across Teams of Knowledge Workers Fabian J. Sting, University of Cologne, C/O Wiso-Sekretariat Universitaetsstrasse Albertus, Köln, 50931, Germany, Matthias Heinz, Johannes Schleef How does routine communication within and across teams of knowledge workers affect their problem solving quality? Our study is based on a randomized controlled trial at a kitchen manufacturer, that is, in the context of complex, mass customized products. Here, knowledge workers virtually meet and discuss quality improvements with frontline colleagues of their team or with other teams in online quality circles. We measure effects on individual quality performance. 4 - Product Development in Crowdfunding: Theoretical and Empirical Analysis Sidika Tunc Candogan, UCL School of Management, London, United Kingdom, Philipp Benjamin Cornelius, Bilal Gokpinar, Ersin Korpeoglu, Christopher S. Tang Crowdfunding goes beyond raising funds. Entrepreneurs often use crowdfunding to solicit feedback from customers to improve their products. We show, both theoretically and empirically, that as the initial development level increases, the likelihood of product improvement during a campaign at first increases and then decreases. Also, while our theoretical model intuitively predicts that the likelihood of campaign success will always increase with the initial development level, our empirical analysis shows that there is first an increase but then an unexpected decrease. We find that this discrepancy can be explained by feature fatigue, and incorporate this effect into our theoretical model to generate prescriptions. While crowdfunding experts believe that products should be as developed as possible before a campaign, we show that this is not always the best strategy. MD26 CC Room 206A In Person: Submodularity in Mixed-Integer Nonlinear Optimization General Session Chair: Qimeng Yu, Northwestern University, Evanston, IL, 60201, United States 1 - Unifying Submodularity And Sequence Submodularity Alexander Stewart Estes, University of Minnesota, Shoreview, MN, 55126-4807, United States In this work, we unify two results regarding greedy algorithms for maximizing submodular functions. For non-decreasing submodular functions defined on sets, it has been shown that the greedy algorithm achieves a solution whose objective is within 1-1/e of the optimal objective. More recent work has defined a new type of submodularity called sequence submodularity, and shown that greedy algorithm likewise achieves 1-1/e of the optimal objective for this type of problem. In our work, we provide a framework that generalizes the concepts of submodularity and sequence submodularity, and show that greedy algorithms achieve 1-1/e of the optimal objective for any optimization problem within this framework. 2 - An Exact Cutting Plane Method for $k$-submodular Function Maximization Qimeng Yu, Northwestern University, IEM S. C210, Evanston, IL, 60201, United States, Simge Küçükyavuz A natural generalization of submodularity—$k$-submodularity—applies to set functions with $k$ arguments and appears in a wide range of applications, such as infrastructure design, machine learning, and healthcare. We propose valid linear inequalities for the hypograph of any $k$-submodular function, and show that maximizing a $k$-submodular function is equivalent to solving a mixed- integer linear program with exponentially many such inequalities. We design the

first exact algorithm to solve general $k$-submodular maximization problems that is not complete enumeration. Our computational experiments on coupled sensor placement demonstrate the efficacy of our method in constrained nonlinear $k$-submodular maximization problems which admit no compact mixed-integer linear formulations. Our method also significantly outperforms exhaustive search. MD27 CC Room 206B In Person: Recent Advances in Vector Optimization General Session Chair: Ozlem Karsu, Bilkent University, Ankara, 6800, Turkey 1 - A Decision Support System for Surgery Rescheduling Problem (SRP) Sajia Afrin Ema, Graduate Student, New Mexico State University, Las Cruces, NM, United States, Sayed Kaes Maruf Hossain, Hansuk Sohn In this research, we have proposed a multiobjective multi-stage mixed-integer linear program (MILP) mathematical model for the surgery rescheduling problem (SRP). To test the model, five (5) instances of the SRP were randomly generated. FICO Xpress Optimizer was used to solve the MILP model using the data instances. However, the solver was unable to provide an optimal solution for all five (5) instances within a reasonable timeframe. To overcome the issue, we have implemented three (3) variants of the Ant colony optimization (ACO) algorithm. All three ACO algorithms were able to provide optimal/near-optimal solutions for all five (5) instances of the SRP with a reasonable timeframe. 2 - Interactive Algorithms to Solve Biobjective and Triobjective Decision Making Problems Ozlem Karsu, Bilkent University, Bilkent University, Ankara, 6800, Turkey, Tugba Denktas, Firdevs Ulus We propose interactive algorithms to find the most preferred solution of biobjective and triobjective integer programming problems. The algorithms can be used in any setting where the decision-maker (DM) has a general monotone utility function. They divide the image space into boxes and search them by solving Pascoletti-Serafini scalarizations, asking questions to the DM to eliminate boxes whenever possible. We also propose a cone based approach that can be incorporated into both algorithms if the DM has a nondecreasing quasiconcave utility function. We demonstrate the performances of the algorithms and their cone based extensions with computational experiments. The results show that interactive algorithms are useful in terms of solution time compared to algorithms that find the whole Pareto set and that the cone based approach leads to less interaction with the DM. MD28 CC Room 207B Technology Tutorial: Your Guide to Financial Portfolio Optimization with Excel/What’sBest! Technology Tutorial 1 - Your Guide to Financial Portfolio Optimization with Excel/What’s Best! Linus Schrage, LINDO Systems, Inc., Chicago, IL, United States There has been an array of risk management optimization models proposed since Harry Markowitz first introduced the mean-variance model. Learn how easy it is to optimize with different risk metrics in Excel with the help of the What’sBest! add-in. In addition to mean-variance, we will cover:

Semi-variance Mean Absolute Deviation (MAD)

Sharpe Ratio Omega Ratio Sortino Ratio Information Ratio Value-at-Risk Conditional Value-at-Risk Power Utility Function Log Utility/Kelly criterion

and a variety of other benchmark tracking methodsBy the end of the session, you will understand when each method should be applied, the common pitfalls of each approach, and the data preparation issues to be concerned with.

80