Iranian Journal of Operations Research

A modified hybrid three-term conjugate gradient method and its applications in image restoration

Ali Ashrafi

In this paper, a modified hybrid three-term conjugate gradient (CG) method is proposed for solving unconstrained optimization problems. The search direction is a three-term hybrid form of the Hestenes-Stiefel (HS) and Liu–Storey (LS) CG parameters. It is established that the method ensures the sufficient descent property independent of line search techniques. The convergence analysis of the proposed method is carried out under standard assumptions for general functions. Numerical experiments on CUTEr problems and image denoising tasks demonstrate that our method outperforms existing approaches in terms of efficiency, accuracy, and robustness, particularly under high levels of salt-and-pepper noise.

A Multi-Objective Optimization Model for Blockchain-Enabled Smart Supply Chains under Uncertainty: Enhancing Transparency, Cost Efficiency, and Sustainability

H. Razavi

This research proposes a robust multi-objective optimization model for blockchain-enabled smart supply chains under uncertainty. The model integrates forward and reverse logistics while incorporating blockchain transaction efficiency to enhance transparency, traceability, and trust among stakeholders. The objectives include minimizing total costs, reducing carbon emissions, maximizing service levels, and optimizing blockchain-related operations. To address uncertainties in demand and transportation costs, the model employs fuzzy robust optimization techniques, ensuring reliable decision-making. To solve the proposed model, several metaheuristic algorithms, including Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and the newly developed Greedy Man Optimization Algorithm (GMOA) are utilized. Comparative analysis demonstrates the superiority of GMOA in achieving high-quality solutions with lower computational time. The results highlight the model’s practical applicability in designing sustainable, transparent, and efficient supply chains. Sensitivity analyses provide managerial insights, emphasizing the impact of key parameters on total costs and operational performance.

Multi-Objective Mathematical Model for Pharmaceutical Location-Routing Problem with Potential Demand Approach

Amir-Mohammad Golmohammadi

Facility location and routing problems have attracted significant research attention since the 1960s due to their practical relevance and complexity. Efficiently establishing production facilities, optimizing vehicle routes, and implementing effective inventory systems are essential for improving organizational performance. In this study, we propose an integrated location-routing model for the pharmaceutical supply chain, designed to satisfy all retailer demands through an appropriate inventory policy, ensuring no demand is unmet. The proposed mixed-integer mathematical model considers a four-tier supply chain, including manufacturers, distributors, wholesalers, and retailers, with the objective of establishing cost-effective warehouses while fulfilling all demand requirements. Demand uncertainty is addressed using a scenario-based probabilistic approach. The model is solved using GAMS for a small-scale case study. For larger-scale instances, where exact solutions are computationally challenging, a meta-heuristic approach—specifically, a genetic algorithm—is employed to efficiently obtain near-optimal solutions.

Computing the Capacity of Sum-networks with Dependent Sources

Mehdi Ghiyasvand

A sum-network is a directed acyclic network with multiple sources and multiple sinks where each sink demands the sum of the independent information generated at the sources. The coding capacity of sum networks with independent sources has been investigated in Tripathy and Ramamoorthy(2015) and it was proven that the upper bound of the coding capacity of such networks is 1. In this paper, it is shown that the upper bound of the coding capacity of a sum network with dependent sources is greater than 1 which is different from the obtained results in Tripathy and Ramamoorthy(2015).
It is also shown that a non-solvable sum-network with independent sources can be converted to a solvable one when the sources have arbitrary dependencies

Conceptualizing Next Generation Supply Chains: Embracing the Platform Paradigm

Mohammad Milad Ahmadi

This study explores the implementation of virtual platforms in supply chain management, emphasizing online production, procurement, and distribution without traditional factory infrastructures. Using a qualitative descriptive-survey approach with inductive reasoning, the research aims to enhance supply chain performance through advanced digital technologies. Rapid advancements in Information and Communication Technologies such as Internet of Things and Artificial Intelligence challenge conventional models by enabling real-time data exchange, improving forecasting accuracy, and reducing delays. Digital integration facilitates seamless communication among suppliers, manufacturers, distributors, and customers, enhancing coordination and cost efficiency. Semi-structured interviews with industry experts were analyzed through thematic analysis, yielding 139 initial codes refined into 25 categories and 5 key themes. These highlight critical dimensions: Digital Integration, Stakeholders Coordination, Edge Computing, Data Analytics and Agility Management. Advanced analytics, leveraging mathematical models and Intelligence algorithms, provide actionable insights for demand forecasting and inventory optimization, strengthening decision-making. The findings underscore the importance of flexibility and agility in addressing market disruptions, with edge computing and real-time data processing identified as vital for operational resilience. Practical recommendations include deploying simulation tools, developing logistics optimization algorithms, and implementing robust cybersecurity protocols. Overall, virtual platforms offer a transformative approach to supply chain management, improving efficiency, reducing costs, and enhancing competitiveness in dynamic markets.

A Fuzzy-Based Approach for Optimizing Dose Distribution in High-Dose-Rate Prostate Brachytherapy Using Multi-Objective Evolutionary Algorithms

Mohammad Mohammadi

Prostate cancer is the most common cancer in men and the second leading cause of cancer-related death worldwide. Over the years, researchers from various fields, beyond medicine, have sought to expand their understanding of the disease to develop more effective treatments. Treatment planning for high-dose-rate (HDR) brachytherapy involves designing the trajectory of the radiation source to deliver sufficient doses to the target area while minimizing exposure to surrounding organs at risk (OAR) within clinically safe limits. Since the exact tumor volume is not known, the model uses gray numbers instead of tumor volume, which provides more accurate results.
In this study, four powerful multi-objective evolutionary algorithms (MOEAs) NSGA¹-II, PESA²-II, SPEA³-II, and MOPSO⁴ are employed. Instead of yielding a single best solution, these algorithms produce a set of Pareto-optimal solutions, each representing a trade-off where no one solution is definitively better than the rest. However, they demonstrate improved performance compared to other optimization methods. The results show that the MOPSO algorithm performs better than the other three powerful algorithms in terms of solution quality and maintaining diversity among solutions.

Utilizing Fuzzy Linear Programming For Addressing Ecological Decision-Making Under Uncertainty

Narjes Amiri

This article examines and analyzes fuzzy linear programming models and techniques. Since its emergence in the 1970s, fuzzy linear programming has addressed the growing complexity of decision-making problems in the real world that occur in uncertain and dynamic environments. Fuzzy linear programming is based on fuzzy set theory and traditional linear programming theory, covering a wide range of theoretical research and algorithmic advancements. Unlike traditional linear programming, fuzzy linear programming does not have a single model, as fuzziness can manifest in various aspects of the model. This paper focuses on solving fuzzy linear programming problems that include inequality constraints. The suggested method employs Yager's linear fuzzy relation, providing a simple and effective way to manage the complexities associated with fuzzy parameters.

Emotion Recognition in Persian Texts Using an Improved Transformer Model

Elham Askari

Emotion recognition in Persian texts using data mining is a significant area within text analysis. Emotions are typically defined as individuals’ emotional reactions to situations, events, and information. Emotion recognition in text involves identifying and analyzing emotional content across various types of textual data. This paper presents a model for detecting different emotions in Persian texts using an enhanced transfer model. The proposed model comprises an encoder and a decoder, each equipped with a self-attention mechanism and RNN modules. Initially, a dataset of sentences annotated with emotional states—anger, happiness, sadness, and fear—is created by multiple users. These sentences are then converted into image representations and fed into the improved transfer model for emotion recognition. Experimental results demonstrate that the model effectively identifies the emotions of sadness, anger, happiness, and surprise with precision, accuracy, recall, and F1-score values of 90.25%, 91.4%, 91.6%, and 90.80%, respectively.

IQ estimation from MRI images using GCNN model

Sara Motamed

Intelligence has long been an interesting and important topic in psychology and cognitive science. IQ is considered a basic measure of a person's cognitive abilities, which includes various aspects of reasoning, problem solving, memory, and overall intellectual ability. Considering the importance of IQ in cognitive and psychological evaluations, the main goal of this article was to provide a new and effective approach to improve the accuracy of estimating this measure through complex brain data processing. In this paper, we have analyzed and developed a hybrid model of GWO algorithm and CNN (GCNN) in order to estimate IQ using brain MRI images. The results of the experiments showed that the accuracy of the proposed model was significantly better than the traditional techniques, and this indicates the high capabilities of the model in interpreting complex medical data. By examining the results, we find that the accuracy of the proposed model with an estimation rate of 93.10% is better than other competing methods.