Modern Research in Decision Making

Abstract The primary objective of this research is to propose an expert system for empowering the sales network in the insurance industry using a fuzzy inference system. This study is applied in nature and employs a descriptive-survey methodology. The statistical population consists of insurance industry experts, managers, sales specialists, and agency affairs professionals. In this study, a range of factors influencing the empowerment of insurance agents was first identified and presented to experts. Following in-depth interviews, five main factors were ultimately selected. Structural equation modeling was used to determine the relationships among the identified factors. The Cronbach's alpha coefficients for all factors ranged between 0.876 and 0.963, indicating acceptable reliability of the measurement. Additionally, the correlations among the research variables were assessed using Pearson’s correlation test. Finally, after designing the conceptual model of the research, each factor was evaluated within the proposed conceptual framework using a fuzzy inference system (FIS) implemented in MATLAB. The proposed system was also applied to three groups of insurance agents. The results of this study demonstrate that the proposed model has a high capability in assessing the key factors of insurance agent empowerment and can effectively support decision-making processes in insurance sales network management. The system’s flexibility and interpretability make it a practical tool for improving the performance of both agents and insurance companies, enabling the implementation of empowerment strategies at both operational and strategic levels.

Abstract The dairy industry faces increasing challenges in maintaining and enhancing the competitiveness of manufacturing firms due to rapid market changes, intensified competition, and shifts in consumer preferences. The main problem addressed in this study is the lack of a dynamic and integrated marketing system capable of analyzing the complex relationships among market variables and proposing strategies aligned with changing conditions. The purpose of this paper is to design a dynamic system of marketing strategies aimed at improving the competitiveness of dairy manufacturing firms. This research adopts a system dynamics approach. Accordingly, key variables were first identified through a review of the literature and expert opinions. Subsequently, causal relationships among the variables were mapped, and the dynamic structure of the system was modeled using stock-and-flow diagrams. In the next stage, plausible scenarios were developed and tested, leading to the formulation and evaluation of six marketing strategies designed to enhance the competitiveness of dairy firms. The results indicate that implementing a dynamic marketing system enables more accurate market data analysis, improves strategic decision-making, and facilitates faster and more effective responses to changes in the competitive environment. The main contribution of this study lies in proposing a dynamic framework based on system dynamics modeling for the design, evaluation, and testing of marketing strategies in the dairy industry.

Abstract The disparity between the official exchange rate and the market rate has created opportunities for currency manipulators to exploit the system. On the other hand, importers of goods are evaluated by the national banking network regardless of their past foreign exchange and financial performance, and collateral is required from them in the form of domestic currency. The main objective of this study is to design an intelligent deep learning model for risk assessment and collateral determination for importers, in such a way that the final model, with the highest level of accuracy and precision, is capable of analyzing large-scale performance data. For this purpose, operational data were first clustered using the K-means method, and the results of the clustering were then used as input for classification models including Random Forest, XGBoost, and Keras Sequential neural networks. The performance of each model was evaluated using the F-measure index. Finally, the combined K-means–RNN model was selected as the best-performing model with the highest accuracy and precision. Using this model, customers were classified into three risk categories, and an optimal mix of cash and non-cash collateral was determined for each group.
The findings indicate that the proposed model is capable of effectively classifying customers based on their performance history and can serve as a robust tool for credit and foreign exchange risk management in the banking sector.

Abstract Inventory management of seasonal products with uncertain quality involves challenging decisions under simultaneous demand and quality uncertainty, while most existing approaches, after inspection, still rely on a rigid binary “accept-or-return” rule. This study develops a three-state threshold newsvendor model for a seasonal perishable product, in which inspection errors, customer returns, logistical limits on replacement, and two-channel discount-sensitive demand are incorporated in an integrated way, and the supplier’s participation constraint is explicitly enforced in designing the optimal policy. Owing to nonlinearity and the presence of min/max operators, the expected profit function is not analytically tractable; therefore, a combined simulation–optimization framework is adopted, where expected profit is estimated via Monte Carlo simulation and the optimal policy is obtained using Bayesian optimization and benchmarked against standard metaheuristics. Numerical results show that, relative to the accept/return policy, activating the replacement mechanism within the three-state framework increases the retailer’s expected profit by about 14.3% and the supplier’s profit by about 6.2%, while simultaneously reducing shortages and waste significantly. Scenario analysis under critical conditions indicates that model performance is robust to parameter changes and that managed replacement acts as an effective risk-hedging instrument. Sensitivity analysis further reveals that, in addition to purchase and selling prices, operational variables such as replacement effectiveness and inspection accuracy have a direct impact on shortages and waste. These findings suggest that moving from traditional full-return contracts toward “smart” contracts embedding a capped replacement mechanism can substantially improve both profitability and operational performance in seasonal perishable supply chains.

Abstract This paper aims to present a model for predicting carbon emissions by considering the role of suppliers, which can significantly contribute to reducing carbon dioxide in the region. To determine the influential variables, literature review and expert opinions (20 experts) were used through the fuzzy Delphi method, ultimately selecting 8 influential variables. For modeling, the Light Gradient Boosting (LGB) algorithm was chosen due to its ability to capture nonlinear dependencies, and the Jellyfish Search Optimizer (JSO) was employed for precise hyperparameter tuning. The research innovation lies in two areas: the integration of LGB and JSO to increase prediction accuracy, and simultaneously considering the role of suppliers as influential variables in carbon emission prediction. The models were implemented using 5,997 data points from Tehran's chemical industries (Khavaran region) during 2022-2025, utilizing Python programming language. Results showed that the hybrid LGB_JSO model, with R² values of 0.9918, 0.9538, and 0.9606 in training, validation, and testing phases respectively, performed better than other models. Temperature, pressure, and storage time were identified as the most important influential parameters.

Abstract In the Iranian pharmaceutical industry, distributors widely rely on trade credit as the primary demand-management instrument to preserve market share. However, unplanned and excessive extensions of credit periods have become a major source of pharmacy liquidity crises, sharp reductions in order volumes, occasional bankruptcies, and, ultimately, diminished patient access to essential medicines. This study develops an optimal trade-credit coordination model and demonstrates that scientifically calibrating the credit duration, using this prevalent industry practice, can substantially alleviate financial distress. The model is formulated within a Stackelberg leader–follower framework and incorporates temperature-dependent Weibull deterioration to capture the cold-chain constraints of pharmaceutical products. Concave fractional programming is employed to establish that the annual total profit function is strictly pseudo-concave, thereby guaranteeing the existence and uniqueness of a global optimal solution for feasible contract parameters. The solution methodology combines an efficient iterative search algorithm with exhaustive brute-force validation implemented in Python. The proposed approach requires no additional financial resources or regulatory changes and can be readily implemented using existing industry practices. Numerical results indicate profit improvements of up to 19.7% and more than a threefold increase in order quantity relative to the decentralized baseline. Overall, the framework provides distributors and pharmacies (retailers) with a practical, cost-free means of transforming a common demand-management instrument into an effective coordination mechanism that enhances profitability, cash-flow stability, and patient access to medicines.