https://www.joee.ir/ Journal of Intelligent Financial Management en daily 1 Tue, 21 Apr 2026 00:00:00 +0330 Tue, 21 Apr 2026 00:00:00 +0330 https://www.joee.ir/article_245465.html The rapid growth of financial technologies (FinTech) and the increasing accessibility of financial markets to retail investors have paved the way for the development of Robo-Advisor systems. By leveraging artificial intelligence algorithms and financial data analytics, these systems provide portfolio management services at lower costs and higher speeds compared to traditional human financial advisors. Despite significant advancements in the field of Robo-Advisors, most existing systems are designed based on classical financial theories and assume that investors behave in a fully rational manner. However, behavioral finance studies have demonstrated that investors'' decisions are influenced by numerous cognitive and emotional biases, which may lead to deviations from rational decision-making.The primary objective of this research is to develop a novel framework for designing a personalized Robo-Advisor that incorporates not only investors’ financial characteristics and risk tolerance but also their behavioral biases into the decision-making process. In this framework, Deep Reinforcement Learning (DRL) serves as the core decision-making mechanism, enabling the intelligent agent to learn optimal asset allocation strategies through continuous interaction with the market environment and analysis of user behavior. By integrating market data, investor-specific characteristics, and behavioral indicators, the proposed model can generate investment recommendations tailored to the unique profile of each investor.The findings of the theoretical analysis suggest that integrating behavioral finance concepts with deep reinforcement learning algorithms can enhance the performance of Robo-Advisors in portfolio management, risk control, and investor satisfaction. Furthermore, the proposed system is expected to mitigate the adverse effects of behavioral biases such as overconfidence, loss aversion, herd behavior, and anchoring, thereby facilitating more rational and effective investment decisions. https://www.joee.ir/article_245469.html In recent years, the evaluation of corporate sustainability performance has become one of the primary concerns of investors, regulatory bodies, and capital markets. The growing emphasis on Environmental, Social, and Governance (ESG) criteria in investment decisions has increased the need for the development of accurate and intelligent models for assessing corporate sustainability. However, most traditional sustainability scoring models are primarily based on the analysis of isolated indicators and structured data, and they are often incapable of capturing the complex relationships among companies, industries, and stakeholders. This limitation reduces assessment accuracy and weakens the ability to predict future sustainability performance.The aim of this study is to develop an intelligent sustainability scoring model for publicly listed companies using Graph Neural Networks (GNNs). In the proposed framework, companies are represented as nodes in a graph, while financial, industrial, ownership, and informational relationships among them are modeled as edges. By employing graph deep learning algorithms, hidden patterns and structural dependencies among companies are extracted and incorporated into the sustainability scoring process.This research adopts an applied and developmental approach within a quantitative research framework. The required data include financial indicators, ESG metrics, ownership information, and inter-company relationship data collected from listed companies. After data preprocessing, the Graph Neural Network model is developed and compared with conventional machine learning techniques, including Random Forest, XGBoost, and Multilayer Perceptron (MLP) neural networks. Model performance is evaluated using metrics such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Accuracy, and the Coefficient of Determination (R²).The findings indicate that incorporating graph structures and inter-company relationships can significantly improve the accuracy and explanatory power of sustainability scoring models. Furthermore, the proposed model provides an effective decision-support tool for investors, corporate managers, and capital market policymakers in promoting sustainability-oriented decision-making and sustainable development. https://www.joee.ir/article_245472.html The transition to a low-carbon economy has introduced “decarbonized assets” as a new financial class. Managing portfolios of these assets is challenging due to geopolitical uncertainty, climate policy volatility, and the dynamic nature of energy and carbon markets. These complexities require adaptive and data-driven optimization approaches capable of handling non-stationary environments.This paper proposes a dynamic multi-objective optimization framework combining metaheuristic algorithms (NSGA-II, PSO, Differential Evolution) with deep recurrent neural networks (LSTM/GRU) for forecasting and decision-making. The model is designed to capture both temporal dependencies and stochastic market behavior. It simultaneously optimizes portfolio return, carbon intensity, and geopolitical risk exposure.Results indicate that the proposed hybrid approach provides more stable and efficient portfolio solutions compared to the classical Markowitz framework, particularly in highly volatile and uncertain market conditions. The findings highlight the importance of integrating artificial intelligence with evolutionary optimization for sustainable financial decision-making. https://www.joee.ir/article_245479.html Technological advancements in the field of quantum computing and the emergence of hybrid quantum–classical algorithms have created new opportunities for addressing complex financial and investment-related problems. One of the most significant challenges in the evaluation of smart infrastructure projects is the presence of multidimensional uncertainties associated with investment costs, market demand, interest rates, technological changes, and governmental policies. Traditional valuation methods, such as Discounted Cash Flow (DCF) analysis, are unable to fully capture the value of strategic opportunities embedded in these projects due to their assumption of managerial inflexibility. In contrast, Real Options Theory provides a more appropriate framework for analyzing investment decisions under uncertainty by explicitly incorporating managerial flexibility into the valuation process.The objective of this study is to develop a novel framework for pricing real options in smart infrastructure projects through the integration of hybrid quantum–classical algorithms and advanced Monte Carlo simulation techniques. Within the proposed framework, the stochastic variables affecting project value are first modeled to represent the underlying sources of uncertainty. Subsequently, potential project value trajectories are generated using multilevel and quasi-random Monte Carlo methods. The theoretical findings indicate that the integration of quantum computing techniques with advanced simulation approaches can significantly improve valuation accuracy while substantially reducing computational costs compared to conventional classical methods. Furthermore, the proposed framework is applicable to a wide range of domains, including smart city projects, smart energy networks, intelligent transportation systems, and digital infrastructure developments. By establishing a connection between engineering economics, computational finance, and quantum computing, this research provides a new pathway for investment decision-making in complex and dynamic environments.