Business research practice is witnessing a surge in the integration of predictive modeling and prescriptive analysis. We describe a modeling framework JANOS that seamlessly integrates the two streams of analytics, allowing researchers and practitioners to embed machine learning models in an end-to-end optimization framework. JANOS allows for specifying a prescriptive model using standard optimization modeling elements such as constraints and variables. The key novelty lies in providing modeling constructs that enable the specification of commonly used predictive models within an optimization model, have the features of the predictive model as variables in the optimization model, and incorporate the output of the predictive models as part of the objective. The framework considers two sets of decision variables: regular and predicted. The relationship between the regular and the predicted variables is specified by the user as pretrained predictive models. JANOS currently supports linear regression, logistic regression, and neural network with rectified linear activation functions. In this paper, we demonstrate the flexibility of the framework through an example on scholarship allocation in a student enrollment problem and provide a numeric performance evaluation.Summary of Contribution. This paper describes a new software tool, JANOS, that integrates predictive modeling and discrete optimization to assist decision making. Specifically, the proposed solver takes as input user-specified pretrained predictive models and formulates optimization models directly over those predictive models by embedding them within an optimization model through linear transformations.
Quantum annealing (QA) can be used to quickly obtain near-optimal solutions for quadratic unconstrained binary optimization (QUBO) problems. In QA hardware, each decision variable of a QUBO should be mapped to one or more adjacent qubits in such a way that pairs of variables defining a quadratic term in the objective function are mapped to some pair of adjacent qubits. However, qubits have limited connectivity in existing QA hardware. This has spurred work on preprocessing algorithms for embedding the graph representing problem variables with quadratic terms into the hardware graph representing qubits adjacencies, such as the Chimera graph in hardware produced by D-Wave Systems. In this paper, we use integer linear programming to search for an embedding of the problem graph into certain classes of minors of the Chimera graph, which we call template embeddings. One of these classes corresponds to complete bipartite graphs, for which we show the limitation of the existing approach based on minimum odd cycle transversals (OCTs). One of the formulations presented is exact and thus can be used to certify the absence of a minor embedding using that template. On an extensive test set consisting of random graphs from five different classes of varying size and sparsity, we can embed more graphs than a state-of-the-art OCT-based approach, our approach scales better with the hardware size, and the runtime is generally orders of magnitude smaller. Summary of Contribution: Our work combines classical and quantum computing for operations research by showing that integer linear programming can be successfully used as a preprocessing step for adiabatic quantum optimization. We use it to determine how a quadratic unconstrained binary optimization problem can be solved by a quantum annealer in which the qubits are coupled as in a Chimera graph, such as in the quantum annealers currently produced by D-Wave Systems. The paper also provides a timely introduction to adiabatic quantum computing and related work on minor embeddings.
Appointment scheduling (AS) plays a crucial role in outpatient clinic management. Traditional methods involve patient grouping using pre-defined rules and scheduling based on these groups. However, pre-defined rules may not adequately capture the heterogeneity in patients’ service times (i.e., consultation duration). Advanced machine learning (ML) methods can address individual-level heterogeneity but pose challenges for practical scheduling. To strike a balance, we propose a data-driven AS decision support system, Cluster-Predict-Schedule (CPS), integrating both supervised and unsupervised ML for efficient patient grouping and scheduling. The novelty of CPS lies in its adaptability to service time heterogeneity through a data-driven approach, determining patient groups based on data rather than pre-defined rules. Additionally, CPS includes a generic and efficient algorithm for generating appointment templates adaptable to any number of patient groups. Our system’s efficacy is demonstrated using a real-world dataset. Evaluated by the weighted sum of patient wait times, physician idle time, and overtime, CPS achieves up to 15.0% cost reduction compared to the FCFA (first-call, first-appointment) scheme and over 4.7% savings against the common New/Return classification with traditional sequencing candidate (TSC) rules. In addition, CPS enhances outpatient operational efficiency without compromising fairness.
Outpatient clinics’ productivity largely depends on their appointment scheduling systems. It is crucial for appointment scheduling to understand the intrinsic heterogeneity in patient and service types and act accordingly. This article describes an outpatient clinic dataset of consultation service time with heterogeneous characteristics. The dataset contains 6637 consultation records collected from 381 half-day sessions between 2018 and 2019. Each record includes encrypted session and patient IDs, consultation start and (approximated) end times, the month and day of the week, whether it was on a holiday, the patient’s visit count for a specific medical condition, gender, whether the consultation was cancer-related, and the distance from the patient’s mailing address to the clinic. These features can be used to classify patients into heterogeneous groups in studies of appointment scheduling. Therefore, this dataset with rich, heterogeneous patient characteristics provides a valuable opportunity for healthcare operations management researchers to develop, test, and benchmark the performance of their models and methods. It can also be used for studying appointment scheduling in other service industries. More generally, it provides pedagogical value in areas related to management science and operations research, applied statistics, and machine learning.
The software and data in this repository are a snapshot of the software and data from the development repository that were used in the research reported on in the paper JANOS: An Integrated Predictive and Prescriptive Modeling Framework by D. Bergman, T. Huang, P. Brooks, A. Lodi, and A.U. Raghunathan.
The rapid adoption of Robotic Mobile Fulfillment Systems (RMFSs) is transforming e-commerce logistics by combining robotic efficiency with human flexibility. However, substantial workload differences between picking stations present an operational challenge in these human–robot collaborative environments. Such imbalance may cause congestion, worker fatigue, and reduced system performance. This paper addresses this efficiency–equity trade-off by evaluating proactive storage planning alongside real-time operational control. We study a joint stowage and storage assignment problem that determines how SKUs should be allocated to racks and where racks should be positioned in the warehouse in order to simultaneously reduce robot travel distances and balance workloads across picking stations. To solve this problem, we develop a data-driven framework that exploits SKU correlation information to cluster frequently co-purchased items and strategically position high-demand racks within the warehouse. Using large-scale operational data from a major e-commerce retailer in China, numerical experiments demonstrate that the combined storage-and-control framework significantly improves system performance. Compared with existing benchmark approaches, the proposed method reduces robot travel distances by 23%, increases system throughput by 27%, and improves workload balance by 62.16%. These results highlight the importance of incorporating workload considerations into storage planning alongside operational control and provide practical guidance for managers seeking to design more efficient and human-centered RMFS operations under the tested operating conditions.