Berk Görgülü
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Berk Görgülü has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
- Vahid Sarhangian
Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, Canada
5 shared publications Jacky Chan
Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, Ontario M5S 3G8, Canada
3 shared publications- Douglas R. Bish
Culverhouse College of Business, The University of Alabama, Tuscaloosa, Alabama 35487
2 shared publications - Ebru K. Bish
Culverhouse College of Business, The University of Alabama, Tuscaloosa, Alabama 35487
2 shared publications - Erhan Can Ozcan
Boston University
2 shared publications - Jing Dong
Columbia University, New York, New York 10027
2 shared publications
Latest Publications
- 1 November 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 74 (6) , 2999-3018
A Novel Adaptive Testing Scheme for Multidisease Testing
Testing Smarter, Not More: An Adaptive Approach In the paper titled “A Novel Adaptive Testing Scheme for Multidisease Testing,” Douglas Bish, Ebru Bish, Merve Bodur, and Berk Görgülü develop an adaptive, cost-efficient testing scheme for lab-based multidisease testing, essential for diagnosis and screening, particularly for disease groups that share symptoms but require disease-specific treatments (e.g., respiratory, gastrointestinal, or sexually transmitted diseases). Their proposed adaptive strategy integrates multiplex assays and pooled testing in a novel way, utilizing modern automated testing technologies to build a customized follow-up test for each person, covering only the diseases flagged in that person’s relevant pooled tests. Optimizing such adaptive designs is mathematically demanding, but in a case study of 18 respiratory diseases, based on data from the Centers for Disease Control and Prevention, their adaptive designs yield substantial cost savings (51%–61%) relative to current practice under forecast error and highly seasonal prevalences. By lowering the cost per subject, adaptive designs can expand testing capacity and support earlier detection under resource constraints.
- 1 November 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 74 (6) , 2999-3018
A Novel Adaptive Testing Scheme for Multidisease Testing
Testing Smarter, Not More: An Adaptive Approach In the paper titled “A Novel Adaptive Testing Scheme for Multidisease Testing,” Douglas Bish, Ebru Bish, Merve Bodur, and Berk Görgülü develop an adaptive, cost-efficient testing scheme for lab-based multidisease testing, essential for diagnosis and screening, particularly for disease groups that share symptoms but require disease-specific treatments (e.g., respiratory, gastrointestinal, or sexually transmitted diseases). Their proposed adaptive strategy integrates multiplex assays and pooled testing in a novel way, utilizing modern automated testing technologies to build a customized follow-up test for each person, covering only the diseases flagged in that person’s relevant pooled tests. Optimizing such adaptive designs is mathematically demanding, but in a case study of 18 respiratory diseases, based on data from the Centers for Disease Control and Prevention, their adaptive designs yield substantial cost savings (51%–61%) relative to current practice under forecast error and highly seasonal prevalences. By lowering the cost per subject, adaptive designs can expand testing capacity and support earlier detection under resource constraints.
- 1 November 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 74 (6) , 2975-2998
Behaviorally Informed Strategies for Infectious Disease Mitigation: Balancing Incentives and Screening
When Screening and Vaccine Incentives Work Better Together When vaccine mandates are not feasible, institutions may rely on vaccination rewards or routine screening of the unvaccinated. The study shows these tools can be more powerful when designed together. Exempting vaccinated individuals turns screening into a nonmonetary incentive while retaining its role in detecting infections. The authors develop a behaviorally informed framework that accounts for vaccine hesitancy, community preferences, limited budgets, and testing capacity. Rather than committing to a rigid plan, the framework learns from surveys and vaccination data and adapts screening as community behavior unfolds. The analysis reveals that incentives and screening are not always substitutes: under some conditions, each strengthens the value of the other. In a realistic college-campus case study using a detailed epidemic model, the adaptive strategy reduced infections and hospitalizations while adhering more closely to budget than incentive-only or screening-only approaches. The findings offer practical guidance for designing flexible mitigation plans for future outbreaks in universities and other close communities.
- 1 November 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 74 (6) , 2975-2998
Behaviorally Informed Strategies for Infectious Disease Mitigation: Balancing Incentives and Screening
When Screening and Vaccine Incentives Work Better Together When vaccine mandates are not feasible, institutions may rely on vaccination rewards or routine screening of the unvaccinated. The study shows these tools can be more powerful when designed together. Exempting vaccinated individuals turns screening into a nonmonetary incentive while retaining its role in detecting infections. The authors develop a behaviorally informed framework that accounts for vaccine hesitancy, community preferences, limited budgets, and testing capacity. Rather than committing to a rigid plan, the framework learns from surveys and vaccination data and adapts screening as community behavior unfolds. The analysis reveals that incentives and screening are not always substitutes: under some conditions, each strengthens the value of the other. In a realistic college-campus case study using a detailed epidemic model, the adaptive strategy reduced infections and hospitalizations while adhering more closely to budget than incentive-only or screening-only approaches. The findings offer practical guidance for designing flexible mitigation plans for future outbreaks in universities and other close communities.
- 15 July 2026
- Published by Springer Nature in Health Care Management Science
- Vol. 29 (3)
Improving hard-to-place kidney allocation: A machine learning approach to center ranking
Kidney transplantation is the preferred treatment for end-stage renal disease, yet donor scarcity and inefficiencies in allocation systems create major bottlenecks, resulting in prolonged wait times and alarming mortality rates. Despite the severe shortage of donor kidneys, timely and effective interventions to prevent non-utilization of life-saving organs remain limited. Expedited out-of-sequence placement of hard-to-place kidneys to centers with a high likelihood of acceptance has been recommended in the literature as a strategy to improve placement success. However, in practice, this process remains nonstandardized and relies heavily on the subjective judgment of decision-makers. We propose a data-driven, machine learning-based ranking policy for out-of-sequence allocation of hard-to-place kidneys that prioritizes transplant centers using predicted center-level acceptance probabilities. Using national deceased-donor and kidney-offer data, we construct a unique offer-level dataset with donor- and center-specific features. We also employ machine learning interpretability tools to provide insight into the factors influencing kidney allocation decisions. Our analysis demonstrates that the proposed policy can reduce the average number of centers considered before placement by fourfold for all kidneys and tenfold for the subset of hard-to-place kidneys. These results highlight the potential of the proposed framework to improve the efficiency of expedited placement and support more timely utilization of hard-to-place kidneys.
- 15 July 2026
- Published by Springer Nature in Health Care Management Science
- Vol. 29 (3)
Improving hard-to-place kidney allocation: A machine learning approach to center ranking
Kidney transplantation is the preferred treatment for end-stage renal disease, yet donor scarcity and inefficiencies in allocation systems create major bottlenecks, resulting in prolonged wait times and alarming mortality rates. Despite the severe shortage of donor kidneys, timely and effective interventions to prevent non-utilization of life-saving organs remain limited. Expedited out-of-sequence placement of hard-to-place kidneys to centers with a high likelihood of acceptance has been recommended in the literature as a strategy to improve placement success. However, in practice, this process remains nonstandardized and relies heavily on the subjective judgment of decision-makers. We propose a data-driven, machine learning-based ranking policy for out-of-sequence allocation of hard-to-place kidneys that prioritizes transplant centers using predicted center-level acceptance probabilities. Using national deceased-donor and kidney-offer data, we construct a unique offer-level dataset with donor- and center-specific features. We also employ machine learning interpretability tools to provide insight into the factors influencing kidney allocation decisions. Our analysis demonstrates that the proposed policy can reduce the average number of centers considered before placement by fourfold for all kidneys and tenfold for the subset of hard-to-place kidneys. These results highlight the potential of the proposed framework to improve the efficiency of expedited placement and support more timely utilization of hard-to-place kidneys.
- 29 June 2026
- Published by Elsevier in Knowledge-Based Systems
- Vol. 350
Prototype Forest in multiple instance learning
- 29 June 2026
- Published by Elsevier in Knowledge-Based Systems
- Vol. 350
Prototype Forest in multiple instance learning
- 4 February 2026
- Published by Springer Nature in Health Care Management Science
- Vol. 29 (1) , 4
Interpretable machine learning for personalized breast cancer screening recommendations
Breast cancer is the most common non-skin cancer and the second leading cause of cancer death in U.S. women. Early detection and timely intervention are thus critical in reducing breast cancer-related deaths. Existing literature for the design of personalized mammography screening is mainly concerned with modeling the problem as a partially observable Markov decision process, which are computationally difficult to solve. In this study, we propose a machine learning-based approach for identifying the personalized screening recommendations using medical history and associated risk factors for individual patients. We find that machine learning models could provide a high degree of accuracy at drastically reduced computational complexity. Furthermore, once trained to sufficient accuracy, we ascertain explainable insights into machine learning model decisions. These insights yield a set of actionable decision rules that healthcare providers could use to support informed patient screening decisions. Overall, our study showcases the potential of machine learning in providing accurate and actionable recommendations for breast cancer screening.
- 4 February 2026
- Published by Springer Nature in Health Care Management Science
- Vol. 29 (1) , 4
Interpretable machine learning for personalized breast cancer screening recommendations
Breast cancer is the most common non-skin cancer and the second leading cause of cancer death in U.S. women. Early detection and timely intervention are thus critical in reducing breast cancer-related deaths. Existing literature for the design of personalized mammography screening is mainly concerned with modeling the problem as a partially observable Markov decision process, which are computationally difficult to solve. In this study, we propose a machine learning-based approach for identifying the personalized screening recommendations using medical history and associated risk factors for individual patients. We find that machine learning models could provide a high degree of accuracy at drastically reduced computational complexity. Furthermore, once trained to sufficient accuracy, we ascertain explainable insights into machine learning model decisions. These insights yield a set of actionable decision rules that healthcare providers could use to support informed patient screening decisions. Overall, our study showcases the potential of machine learning in providing accurate and actionable recommendations for breast cancer screening.
- Open Access
- 18 September 2025
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 8, 1662960
Predicting offer burden to optimize batch sizes in simultaneously expiring kidney offers
Background: Timely and efficient allocation of deceased donor kidneys is a persistent challenge in transplantation. Traditional sequential offer systems often lead to extended delays and high nonuse rates, as many kidneys undergo multiple refusals before being accepted. Simultaneously expiring offers, where a kidney is offered to a batch of centers with synchronized response deadlines, offer a more efficient alternative. However, fixed batch sizes fail to account for variability in offer requirements, potentially introducing new inefficiencies or overwhelming transplant professionals with excessive notifications.Methods: We investigated the use of machine learning-based survival models to dynamically predict the number of offers a kidney will require before acceptance. Utilizing data on over 16,000 deceased donor kidneys from the national organ offer dataset, we engineered predictive features from both donor profiles and recipient pool characteristics. We trained and evaluated multiple survival models using time-dependent concordance indices along with other survival and regression performance metrics.Results: The Random Survival Forest model achieved the best performance, with a time-dependent C-index of 0.882, effectively estimating the required offer volume for kidney placement. Feature importance analysis revealed that waitlist characteristics, such as mean Estimated Post-Transplant Survival (EPTS), mean Calculated Panel Reactive Antibody (CPRA), time on dialysis, and waitlist duration, were among the most influential predictors. When integrated into a dynamic simultaneous offer system, these predictions have the potential to reduce average placement delays from 17.37 h to 1.59 h while maintaining a manageable level of extraneous offers.Discussion: Our results demonstrate that survival-based predictive modeling can meaningfully improve the efficiency of simultaneously expiring offers in kidney allocation. By personalizing batch sizes based on expected offer burden, such models can reduce delays without overwhelming transplant professionals. These findings underscore the value of integrating real-time, data-driven tools into organ allocation systems to improve operational efficiency and facilitate practical implementation.
- 18 September 2025
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 8, 1662960
Open AccessPredicting offer burden to optimize batch sizes in simultaneously expiring kidney offers
Background: Timely and efficient allocation of deceased donor kidneys is a persistent challenge in transplantation. Traditional sequential offer systems often lead to extended delays and high nonuse rates, as many kidneys undergo multiple refusals before being accepted. Simultaneously expiring offers, where a kidney is offered to a batch of centers with synchronized response deadlines, offer a more efficient alternative. However, fixed batch sizes fail to account for variability in offer requirements, potentially introducing new inefficiencies or overwhelming transplant professionals with excessive notifications.Methods: We investigated the use of machine learning-based survival models to dynamically predict the number of offers a kidney will require before acceptance. Utilizing data on over 16,000 deceased donor kidneys from the national organ offer dataset, we engineered predictive features from both donor profiles and recipient pool characteristics. We trained and evaluated multiple survival models using time-dependent concordance indices along with other survival and regression performance metrics.Results: The Random Survival Forest model achieved the best performance, with a time-dependent C-index of 0.882, effectively estimating the required offer volume for kidney placement. Feature importance analysis revealed that waitlist characteristics, such as mean Estimated Post-Transplant Survival (EPTS), mean Calculated Panel Reactive Antibody (CPRA), time on dialysis, and waitlist duration, were among the most influential predictors. When integrated into a dynamic simultaneous offer system, these predictions have the potential to reduce average placement delays from 17.37 h to 1.59 h while maintaining a manageable level of extraneous offers.Discussion: Our results demonstrate that survival-based predictive modeling can meaningfully improve the efficiency of simultaneously expiring offers in kidney allocation. By personalizing batch sizes based on expected offer burden, such models can reduce delays without overwhelming transplant professionals. These findings underscore the value of integrating real-time, data-driven tools into organ allocation systems to improve operational efficiency and facilitate practical implementation.
- Open Access
Multiclass Queue Scheduling Under Slowdown: An Approximate Dynamic Programming Approach
In many service systems, especially those in healthcare, customer waiting times can result in increased service requirements. Such service slowdowns can significantly impact system performance. Therefore, it is important to properly account for their impact when designing scheduling policies. Scheduling under wait-dependent service times is challenging, especially when multiple customer classes are heterogeneously affected by waiting. In this work, we study scheduling policies in multiclass, multiserver queues with wait-dependent service slowdowns. We propose a simulation-based Approximate Dynamic Programming (ADP) algorithm to find close-to-optimal scheduling policies. The ADP algorithm (i) represents the policy using classifiers based on the index policy structure, (ii) leverages a coupling method to estimate the differences of the relative value functions directly, and (iii) uses adaptive sampling for efficient state-space exploration. Through extensive numerical experiments, we illustrate that the ADP algorithm generates close-to-optimal policies that outperform well-known benchmarks. We also provide insights into the structure of the optimal policy, which reveals an important trade-off between instantaneous cost reduction and preventing the system from reaching high-cost equilibria. Lastly, we conduct a case study on scheduling admissions into rehabilitation care to illustrate the effectiveness of the ADP algorithm in practice.
Open AccessMulticlass Queue Scheduling Under Slowdown: An Approximate Dynamic Programming Approach
In many service systems, especially those in healthcare, customer waiting times can result in increased service requirements. Such service slowdowns can significantly impact system performance. Therefore, it is important to properly account for their impact when designing scheduling policies. Scheduling under wait-dependent service times is challenging, especially when multiple customer classes are heterogeneously affected by waiting. In this work, we study scheduling policies in multiclass, multiserver queues with wait-dependent service slowdowns. We propose a simulation-based Approximate Dynamic Programming (ADP) algorithm to find close-to-optimal scheduling policies. The ADP algorithm (i) represents the policy using classifiers based on the index policy structure, (ii) leverages a coupling method to estimate the differences of the relative value functions directly, and (iii) uses adaptive sampling for efficient state-space exploration. Through extensive numerical experiments, we illustrate that the ADP algorithm generates close-to-optimal policies that outperform well-known benchmarks. We also provide insights into the structure of the optimal policy, which reveals an important trade-off between instantaneous cost reduction and preventing the system from reaching high-cost equilibria. Lastly, we conduct a case study on scheduling admissions into rehabilitation care to illustrate the effectiveness of the ADP algorithm in practice.
- Open Access
Optimizing Hard-to-Place Kidney Allocation: A Machine Learning Approach to Center Ranking
Kidney transplantation is the preferred treatment for end-stage renal disease, yet the scarcity of donors and inefficiencies in allocation systems create major bottlenecks, resulting in prolonged wait times and alarming mortality rates. Despite their severe scarcity, timely and effective interventions to prevent non-utilization of life-saving organs remain inadequate. Expedited out-of-sequence placement of hard-to-place kidneys to centers with the highest likelihood of utilizing them has been recommended in the literature as an effective strategy to improve placement success. Nevertheless, current attempts towards this practice is non-standardized and heavily rely on the subjective judgment of the decision-makers. This paper proposes a novel data-driven, machine learning-based ranking system for allocating hard-to-place kidneys to centers with a higher likelihood of accepting and successfully transplanting them. Using the national deceased donor kidney offer and transplant datasets, we construct a unique dataset with donor-, center-, and patient-specific features. We propose a data-driven out-of-sequence placement policy that utilizes machine learning models to predict the acceptance probability of a given kidney by a set of transplant centers, ranking them accordingly based on their likelihood of acceptance. Our experiments demonstrate that the proposed policy can reduce the average number of centers considered before placement by fourfold for all kidneys and tenfold for hard-to-place kidneys. This significant reduction indicates that our method can improve the utilization of hard-to-place kidneys and accelerate their acceptance, ultimately reducing patient mortality and the risk of graft failure. Further, we utilize machine learning interpretability tools to provide insights into factors influencing the kidney allocation decisions.
Open AccessOptimizing Hard-to-Place Kidney Allocation: A Machine Learning Approach to Center Ranking
Kidney transplantation is the preferred treatment for end-stage renal disease, yet the scarcity of donors and inefficiencies in allocation systems create major bottlenecks, resulting in prolonged wait times and alarming mortality rates. Despite their severe scarcity, timely and effective interventions to prevent non-utilization of life-saving organs remain inadequate. Expedited out-of-sequence placement of hard-to-place kidneys to centers with the highest likelihood of utilizing them has been recommended in the literature as an effective strategy to improve placement success. Nevertheless, current attempts towards this practice is non-standardized and heavily rely on the subjective judgment of the decision-makers. This paper proposes a novel data-driven, machine learning-based ranking system for allocating hard-to-place kidneys to centers with a higher likelihood of accepting and successfully transplanting them. Using the national deceased donor kidney offer and transplant datasets, we construct a unique dataset with donor-, center-, and patient-specific features. We propose a data-driven out-of-sequence placement policy that utilizes machine learning models to predict the acceptance probability of a given kidney by a set of transplant centers, ranking them accordingly based on their likelihood of acceptance. Our experiments demonstrate that the proposed policy can reduce the average number of centers considered before placement by fourfold for all kidneys and tenfold for hard-to-place kidneys. This significant reduction indicates that our method can improve the utilization of hard-to-place kidneys and accelerate their acceptance, ultimately reducing patient mortality and the risk of graft failure. Further, we utilize machine learning interpretability tools to provide insights into factors influencing the kidney allocation decisions.
- Open Access
Smooth Ranking SVM via Cutting-Plane Method
The most popular classification algorithms are designed to maximize classification accuracy during training. However, this strategy may fail in the presence of class imbalance since it is possible to train models with high accuracy by overfitting to the majority class. On the other hand, the Area Under the Curve (AUC) is a widely used metric to compare classification performance of different algorithms when there is a class imbalance, and various approaches focusing on the direct optimization of this metric during training have been proposed. Among them, SVM-based formulations are especially popular as this formulation allows incorporating different regularization strategies easily. In this work, we develop a prototype learning approach that relies on cutting-plane method, similar to Ranking SVM, to maximize AUC. Our algorithm learns simpler models by iteratively introducing cutting planes, thus overfitting is prevented in an unconventional way. Furthermore, it penalizes the changes in the weights at each iteration to avoid large jumps that might be observed in the test performance, thus facilitating a smooth learning process. Based on the experiments conducted on 73 binary classification datasets, our method yields the best test AUC in 25 datasets among its relevant competitors.
Open AccessSmooth Ranking SVM via Cutting-Plane Method
The most popular classification algorithms are designed to maximize classification accuracy during training. However, this strategy may fail in the presence of class imbalance since it is possible to train models with high accuracy by overfitting to the majority class. On the other hand, the Area Under the Curve (AUC) is a widely used metric to compare classification performance of different algorithms when there is a class imbalance, and various approaches focusing on the direct optimization of this metric during training have been proposed. Among them, SVM-based formulations are especially popular as this formulation allows incorporating different regularization strategies easily. In this work, we develop a prototype learning approach that relies on cutting-plane method, similar to Ranking SVM, to maximize AUC. Our algorithm learns simpler models by iteratively introducing cutting planes, thus overfitting is prevented in an unconventional way. Furthermore, it penalizes the changes in the weights at each iteration to avoid large jumps that might be observed in the test performance, thus facilitating a smooth learning process. Based on the experiments conducted on 73 binary classification datasets, our method yields the best test AUC in 25 datasets among its relevant competitors.