Vikrant Vaze

Professor Vikrant Vaze is co-author of "Integrated Schedule Planning for Regional Airlines Using Column Generation" published in Transportation Science. "Where, how often, when, and with what aircraft an airline operates its flights decides passenger demand and costs. Ours is the first study to jointly optimize all these decisions by developing a new algorithm resulting in significant profit improvements and better ability to serve small and rural communities with connections to larger hub airports," said Vaze.

Professor Vikrant Vaze is co-author of "Multimodal Transportation Pricing Alliance Design: Large-Scale Optimization for Rapid Gains" selected by the Transportation Science editorial board to receive one of the journal's Paper of the Year Awards for 2025. "We show that innovative pricing optimization algorithms and mechanism design can incentivize collaboration between public transit agencies and private on-demand operators to simultaneously benefit passengers, increase provider profits, reduce emissions and improve equity—a rare win-win-win-win opportunity," said Vaze.

Professor Vikrant Vaze was invited to speak at the GE Vernova AI EDGE Symposium in New York. His talk, titled "Multi-Objective Stochastic Optimization of Hybrid Renewable Energy Systems in Fragile States," was based on his work done with PhD students Lilly Yang and Siqi Ke and MS student Ryan Proulx '25 as co-authors. "The biggest takeaways from the symposium were about the future of work itself, about the industry-wide opportunities and challenges associated with the rapidly evolving business practices to accommodate AI-human teams—learning what to delegate to AI and what must be human-controlled," said Vaze.

PhD candidate Yineng "Phoebe" Sun Th'17 Th'26 is first author, with Lecturer Armin Fügenschuh and Professor Vikrant Vaze as co-authors, of "Physics-aware truck and drone delivery planning using optimization & machine learning" published in Computers & Operations Research. The presented framework has been shown to reduce delivery time and energy use compared with simplified planning approaches. "This work matters because it brings realistic physical modeling into operations research, showing how optimization and data-driven methods can work together to make complex delivery systems both more efficient and more practical," said Sun.

PhD student Spencer Bertsch Th'19 and Professor Vikrant Vaze are co-authors of "FIND Lp(a) Machine Learning Model: Targeted Screening Enrichment of Elevated Lipoprotein(a) in Atherosclerotic Cardiovascular Disease" published in JACC: Advances. The study used data from deidentified medical records to "Flag, Identify, Network, and Deliver (FIND)" adults with likely elevated Lipoprotein(a). "By combining state-of-the-art machine learning with relevant clinical information from electronic health records, we demonstrated that the effectiveness of Lipoprotein(a) screening can be more than doubled," said Bertsch.

Professors Ryan Halter, Jonathan Elliott, Vikrant Vaze, and Ethan Murphy—with Geisel Professor Norman Paradis—were issued a US patent for "System and method to detect the presence and progression of diseases characterized by systemic changes in the state of the vasculature." The invention uses a novel technique to obtain multiple tissue measurements which are then "transformed by a multivariate algorithm to outputs that convey the diagnostic and prognostic risk of the disease of interest," according to the patent. "We show that by effectively combining signals from multiple sensors using advanced machine learning algorithms, we can save lives through early detection of hidden internal bleeding," said Vaze.

Professor Vikrant Vaze co-authored "Large-Scale Airline Crew Recovery Using Mixed-Integer Optimization and Supervised Machine Learning" published in Transportation Science. Based on work by Vaze's co-advisee at MIT, Ahmet Esat Hizir (pictured), this research won the "Best Innovation" award at AGIFORS' 2024 Crew Management Study Group Meeting. "By teaching a computer to learn from past disruption recovery attempts and then guiding a powerful optimizer with those lessons, we have built a fast, flexible tool that helps airlines get their crews back on schedule more efficiently, cut costs dramatically, and reduce the ripple effects on passengers," said Vaze.

PhD student Mai Pham, will present her paper, coauthored with professors Vikrant Vaze and Peter Chin, titled "Advancing Differentiable Mechanism Design: Neural architectures for combinatorial auctions" for a workshop at the Conference on Neural Information Processing Systems. Although auctions are considered an effective way of allocating limited resources when demand is high, designing auctions that are simultaneously optimal for the participants, system operator, and greater society is challenging. The paper presents a new approach that leverages modern deep learning architectures and algorithms to meet this challenge.

PhD student Mai Pham and professors Vikrant Vaze and Peter Chin coauthored "Strategic Cyber Defense via Reinforcement Learning-Guided Combinatorial Auctions" which was chosen for presentation at the IEEE High Performance Extreme Computing Conference. "This work formulates the cyber defense problem, in which we act as a defense agent protecting a network from malware attackers, as a combinatorial auction," said Pham. "We combined reinforcement learning with differentiable mechanism design to find the best resource allocation for the hosts, which aims to upstream decision making and proactively protect the network."

Professor Vikrant Vaze is a co-author of "A novel outreach approach for identification of familial hypercholesterolemia: Interview-based formative evaluation to improve healthcare access and quality" published in PEC Innovation. "This was a collaborative effort with folks from DH and Geisel, as well as Family Heart Foundation. The study is aimed at designing and evaluating direct outreach and referral to specialty care for patients with an elevated risk of FH identified through a machine learning model and expert review of the electronic record in a rural US health system. It's an excellent human-centered design thinking exercise and it yielded a great deal of success," said Vaze.

Professor Vikrant Vaze is a co-author of “Operational Research: methods and applications” which recently became the most read article of all time in the Journal of the Operational Research Society—the oldest journal in the field of operations research. "This is arguably the first prominent article to provide a comprehensive overview of the state-of-the-art in operations research [OR], from both a methodological and from an applications standpoint. It is meant to serve as the first point of reference for OR academics, researchers, students and practitioners alike," said Vaze.

Professors Vikrant Vaze and Peter Chin are co-authors of "Implementation of a Machine Learning Model and Direct-to-Patient Outreach Program for Targeted Screening for Familial Hypercholesterolemia" published in Journal of Clinical Lipidology. "Machine learning by itself may not make a big difference," says Vaze. "The difference is what you do with the results. Ultimately, the combination of expert knowledge and careful operationalizing of machine learning can make a major practical impact."

PhD student Navid Rashedi (pictured), Professor Ethan Murphy, Alexandra Hamlin '16 Th'17 Th'19, research associate Victor Borza, and Professors Jonathan Elliott, Ryan Halter, and Vikrant Vaze are co-authors of: "Detection of occult hemorrhage using multivariate non-invasive technologies" published in Physiological Measurement. "This work investigated machine learning to combine multiple technologies—electrical impedance and near infrared spectroscopy—to better detect internal bleeds in a porcine study. Internal bleeds are often not detectable until it's too late. This approach appears to detect them earlier and more accurately," said Murphy. 

Professor Vikrant Vaze received the INFORMS AAS Best Paper Award as co-author of, "Vertiport Planning for Urban Aerial Mobility: An Adaptive Discretization Approach" published in Manufacturing & Service Operations Management. The paper optimizes the number, locations, and capacities of vertiports for electric vertical-takeoff-and-landing (eVTOL) vehicles in urban aerial mobility (UAM) systems while capturing interdependencies between vertiport deployment, tactical operations, and passenger demand.

Professor Vikrant Vaze received the "Best Innovation" award at the Airline Group of the International Federation of Operational Research Societies' (AGIFORS) 2024 Crew Management Study Group Meeting in Modena, Italy. Vaze earned the award with a technical presentation titled "Integrated Crew Recovery Using Optimization and Machine Learning." Said Vaze, "This research combines interpretable machine learning with optimization to obtain high-quality airline recovery solutions rapidly and transparently."

PhD students Spencer Bertsch Th'19 and Navid Rashedi, alum Yifei Sun Th'22, and Professors Ethan Murphy (first author), Jonathan Elliott, Ryan Halter, and Vikrant Vaze—along with DHMC and Mayo Clinic researchers—co-authored "Non-invasive biomarkers for detecting progression toward hypovolemic cardiovascular instability in a lower body negative pressure model" published in Scientific Reports. The paper summarizes how electrical impedance tomography (EIT) can be used as a novel marker for early bleed detection.

Professor Vikrant Vaze received the INFORMS AAS Best Paper Award this week at the INFORMS annual meeting in Phoenix. The paper entitled, "Choice-Based Airline Schedule Design and Fleet Assignment: A Decomposition Approach," proposes a "decomposition approach based on partitioning the flight network into smaller subnetworks by exploiting weak dependencies in network structure."

Professors Liz Murnane, Vikrant Vaze, and Wesley Marrero are lead faculty on a team selected as one of 34 semifinalists for the inaugural NSF Regional Innovation Engines competition—spanning nearly all key technology areas and societal and economic challenges highlighted in the "CHIPS and Science Act." A partnership with UMass Lowell, their proposed engine is titled, "Advancing Health Equity Through Digital Technologies, Data Infrastructure, and Artificial Intelligence." Each Engine could receive up to $160 million over 10 years.

Professor Vikrant Vaze is one of only two academics cited in a report from The Government Accountability Office titled, "AIRLINE PASSENGER PROTECTIONS: Observations on Flight Delays and Cancellations, and DOT's Efforts to Address Them." Says Vaze on why his research was cited, "US airlines have faced some unusual disruptions recently. Our research has focused on how these disruptions impact passenger travel, and the importance of measuring the success of any disruption mitigation effort in terms of the extent to which it reduces passenger inconvenience."