Ömer Faruk Yılmaz
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ömer Faruk Yılmaz has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Beren Gürsoy Yılmaz
Karadeniz Technical University
12 shared publications - Fatma Betül Yeni
Karadeniz Technical University
7 shared publications - Gökhan Özçelik
Karadeniz Technical University
7 shared publications - Alperen Bal
American University of the Middle East
2 shared publications - Emre Cevikcan
Istanbul Technical University
2 shared publications Hikmet Erbiyik
Yalova University, Turkey
2 shared publications
Latest Publications
- 18 October 2026
- Published by Taylor & Francis in International Journal of Systems Science: Operations & Logistics
- Vol. 13 (1)
Role of warehouse configuration in supply chain resilience under uncertainty: a stochastic programming approach with data-driven product clustering
1. In today’s uncertain global environment, enhancing supply chain resilience has become essential as disruptions increasingly affect both demand patterns and supplier performance. Recent crises, p...
- 18 October 2026
- Published by Taylor & Francis in International Journal of Systems Science: Operations & Logistics
- Vol. 13 (1)
Role of warehouse configuration in supply chain resilience under uncertainty: a stochastic programming approach with data-driven product clustering
1. In today’s uncertain global environment, enhancing supply chain resilience has become essential as disruptions increasingly affect both demand patterns and supplier performance. Recent crises, p...
- Open Access
- 29 March 2026
- Published by Communications Faculty of Sciences University of Ankara Series A1 Mathematics and Statistics in Communications Faculty of Sciences University of Ankara Series A1 Mathematics and Statistics
- Vol. 75 (1) , 89-112
Two-Stage Stochastic Programming and Robust Optimization Models for Resilient Supply Chain Network Design under Uncertainty: A Real Case Study
This paper investigates the resilient multi-period, multi-stage supply chain (SC) network design problem under demand and raw material quality uncertainty within a just-in-time (JIT) distribution setting, based on a real case study. The proposed approach models a four-stage SC comprising suppliers, manufacturers, distributors, and retailers, and develops two-stage stochastic programming and robust optimization models to enhance resilience. Unlike existing studies, this research uniquely integrates JIT distribution with the simultaneous consideration of demand and raw material quality uncertainties, providing practical, data-driven insights for decision-makers. Computational results show that the proposed models produce applicable solutions for real-world implementation. Across all models, high-quality raw materials are preferred 54% in the deterministic model, 55% on average in 30 of 40 stochastic scenarios, and 52% in the robust model, even under worst-case conditions. These findings indicate that prioritizing high-quality raw materials, despite higher purchasing costs, is crucial for maintaining JIT principles and ensuring on-time deliveries. Furthermore, the results highlight that the strategic location of distributors is critical to meeting retailers’ demand at the right time and in the right quantity.
- 29 March 2026
- Published by Communications Faculty of Sciences University of Ankara Series A1 Mathematics and Statistics in Communications Faculty of Sciences University of Ankara Series A1 Mathematics and Statistics
- Vol. 75 (1) , 89-112
Open AccessTwo-Stage Stochastic Programming and Robust Optimization Models for Resilient Supply Chain Network Design under Uncertainty: A Real Case Study
This paper investigates the resilient multi-period, multi-stage supply chain (SC) network design problem under demand and raw material quality uncertainty within a just-in-time (JIT) distribution setting, based on a real case study. The proposed approach models a four-stage SC comprising suppliers, manufacturers, distributors, and retailers, and develops two-stage stochastic programming and robust optimization models to enhance resilience. Unlike existing studies, this research uniquely integrates JIT distribution with the simultaneous consideration of demand and raw material quality uncertainties, providing practical, data-driven insights for decision-makers. Computational results show that the proposed models produce applicable solutions for real-world implementation. Across all models, high-quality raw materials are preferred 54% in the deterministic model, 55% on average in 30 of 40 stochastic scenarios, and 52% in the robust model, even under worst-case conditions. These findings indicate that prioritizing high-quality raw materials, despite higher purchasing costs, is crucial for maintaining JIT principles and ensuring on-time deliveries. Furthermore, the results highlight that the strategic location of distributors is critical to meeting retailers’ demand at the right time and in the right quantity.
- 24 March 2026
- Published by Elsevier in European Journal of Operational Research
- Vol. 334 (2) , 420-437
A unified robust stochastic programming approach for integrated production and distribution scheduling problem for seru production system under uncertainty
- 24 March 2026
- Published by Elsevier in European Journal of Operational Research
- Vol. 334 (2) , 420-437
A unified robust stochastic programming approach for integrated production and distribution scheduling problem for seru production system under uncertainty
- Open Access
- 24 June 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (13) , 7112
Scalable and Efficient Protein Secondary Structure Prediction Using Autoencoder-Reduced ProtBERT Embeddings
This study proposes a deep learning framework for Protein Secondary Structure Prediction (PSSP) that prioritizes computational efficiency while preserving classification accuracy. Leveraging ProtBERT-derived embeddings, we apply autoencoder-based dimensionality reduction to compress high-dimensional sequence representations. These are segmented into fixed-length subsequences, enabling efficient input formatting for a Bi-LSTM-based classifier. Our experiments, conducted on a curated PISCES-based dataset, reveal that reducing input dimensions from 1024 to 256 preserves over 99% of predictive performance (Q3 F1 score: 0.8049 → 0.8023) while reducing GPU memory usage by 67% and training time by 43%. Moreover, subsequence lengths of 50 residues provide an optimal trade-off between contextual learning and training stability. Compared to baseline configurations, the proposed framework reduces training overhead substantially without compromising structural accuracy in both the Q3 and Q8 classification schemes. These findings offer a practical pathway for scalable protein structure prediction, particularly in resource-constrained environments.
- 24 June 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (13) , 7112
Open AccessScalable and Efficient Protein Secondary Structure Prediction Using Autoencoder-Reduced ProtBERT Embeddings
This study proposes a deep learning framework for Protein Secondary Structure Prediction (PSSP) that prioritizes computational efficiency while preserving classification accuracy. Leveraging ProtBERT-derived embeddings, we apply autoencoder-based dimensionality reduction to compress high-dimensional sequence representations. These are segmented into fixed-length subsequences, enabling efficient input formatting for a Bi-LSTM-based classifier. Our experiments, conducted on a curated PISCES-based dataset, reveal that reducing input dimensions from 1024 to 256 preserves over 99% of predictive performance (Q3 F1 score: 0.8049 → 0.8023) while reducing GPU memory usage by 67% and training time by 43%. Moreover, subsequence lengths of 50 residues provide an optimal trade-off between contextual learning and training stability. Compared to baseline configurations, the proposed framework reduces training overhead substantially without compromising structural accuracy in both the Q3 and Q8 classification schemes. These findings offer a practical pathway for scalable protein structure prediction, particularly in resource-constrained environments.
- Open Access
- 13 May 2025
- Published by Springer Nature in Operational Research
- Vol. 25 (2) , 1-42
A bi-objective robust optimization model to bolster a resilient medical supply chain in case of the ripple effect
The medical supply chain (MSC) plays a crucial role in ensuring that drugs, vaccines, gloves, and other medical products are delivered to the right place at the right times and in the required quantity. For this reason, the viability of MSC is vital in case of an external risk such as the coronavirus (COVID-19) pandemic. The COVID-19 pandemic is a type of ripple effect that has devastating effect on supply chain performance. With that in mind, to the best of our knowledge, this study explores a bi-objective MSC design problem under uncertainty caused by the ripple effect for the first time. Accordingly, a generic bi-objective robust optimization model is fundamentally developed to represent the addressed problem mathematically by considering the uncertainty sourcing from pandemic. To obtain applicable results, a real case study is considered for MSC design in Istanbul/Turkey as a practical contribution by validating the optimization model. Furthermore, a set of scenarios are generated by placing an emphasis on the decrease in capacity utilization rates and the increase in product demand due to the pandemic. A computational study is conducted through scenarios and risk mitigation strategies to reveal managerial insights by combining the strategic and operational level decisions regarding the MSC network. The improved augmented ϵ-constrained (AUGMECON2) method is employed to obtain diversified Pareto-optimal solutions for all problems. Several comparison metrics are employed to further analyze the solutions from different perspectives. According to the computational results attained by extensive experiments, a unified strategy is proposed to achieve MSC resiliency. Besides, solving large sized problem instances through the proposed methodology is highlighted as the main limitation of this study.
- 13 May 2025
- Published by Springer Nature in Operational Research
- Vol. 25 (2) , 1-42
Open AccessA bi-objective robust optimization model to bolster a resilient medical supply chain in case of the ripple effect
The medical supply chain (MSC) plays a crucial role in ensuring that drugs, vaccines, gloves, and other medical products are delivered to the right place at the right times and in the required quantity. For this reason, the viability of MSC is vital in case of an external risk such as the coronavirus (COVID-19) pandemic. The COVID-19 pandemic is a type of ripple effect that has devastating effect on supply chain performance. With that in mind, to the best of our knowledge, this study explores a bi-objective MSC design problem under uncertainty caused by the ripple effect for the first time. Accordingly, a generic bi-objective robust optimization model is fundamentally developed to represent the addressed problem mathematically by considering the uncertainty sourcing from pandemic. To obtain applicable results, a real case study is considered for MSC design in Istanbul/Turkey as a practical contribution by validating the optimization model. Furthermore, a set of scenarios are generated by placing an emphasis on the decrease in capacity utilization rates and the increase in product demand due to the pandemic. A computational study is conducted through scenarios and risk mitigation strategies to reveal managerial insights by combining the strategic and operational level decisions regarding the MSC network. The improved augmented ϵ-constrained (AUGMECON2) method is employed to obtain diversified Pareto-optimal solutions for all problems. Several comparison metrics are employed to further analyze the solutions from different perspectives. According to the computational results attained by extensive experiments, a unified strategy is proposed to achieve MSC resiliency. Besides, solving large sized problem instances through the proposed methodology is highlighted as the main limitation of this study.
- Top Cited Papers
- 4 April 2025
- Published by Elsevier in International Journal of Production Economics
- Vol. 285
Investigating the impact of strategic warehouse design and product clustering on supply chain viability: A unified robust stochastic programming approach
- 4 April 2025
- Published by Elsevier in International Journal of Production Economics
- Vol. 285
Top Cited PapersInvestigating the impact of strategic warehouse design and product clustering on supply chain viability: A unified robust stochastic programming approach
- 2 February 2025
- Published by Springer Nature
- p. 184-195
Supplier Selection for Agriculture Industry Under Uncertainty: Machine Learning Based Sample Average Approximation Method
- 2 February 2025
- Published by Springer Nature
- p. 184-195
Supplier Selection for Agriculture Industry Under Uncertainty: Machine Learning Based Sample Average Approximation Method
- 2 February 2025
- Published by Springer Nature
- p. 127-140
Climate Risk Management for Aquaculture Industry: Robust Programming Approach vs Random Forest Algorithm
- 2 February 2025
- Published by Springer Nature
- p. 127-140
Climate Risk Management for Aquaculture Industry: Robust Programming Approach vs Random Forest Algorithm