Multi-objective Decision-making Problem Research Articles

Credibilistic portfolio optimization with higher-order moments using coherent triangular fuzzy numbers

Financial portfolio formation is usually a multi-objective decision-making problem concerning return and risk on the investment. In this study, we make use of an extension of regular triangular fuzzy numbers, known as coherent triangular fuzzy numbers, to describe portfolio returns in a credibility-based framework. The paper’s novelty lies in proposing and deriving a crisp equivalent for computing the coherent triangular fuzzy number-based credibilistic semivariance, credibilistic skewness, and credibilistic semikurtosis. Utilizing these analytical expressions, along with additional formulations found in existing literature, we present three multi-objective portfolio optimization models involving the practical constraints related to investment decisions. All the proposed analytical expressions, when used with the returns of the portfolio as a whole, help to overcome the computationally expensive process of simulating results using the returns of individual assets. The proposed models differ with respect to different risk measures, viz. semivariance, Mean-Absolute-Semi-Deviation (MASD), and Conditional Value-at-Risk (CVaR). These models are solved using an adaptation of an efficient Multi-Objective Genetic Algorithm (MOGA) specifically designed to solve portfolio optimization problems with practical constraints. Data from the National Stock Exchange (NSE) in Mumbai, India and the New York Stock Exchange (NYSE) in New York, USA, are used to demonstrate the effectiveness of the proposed portfolio optimization models and solution methodology. All the proposed models are compared with respect to each other and the benchmarks considered in this study to bring out the performance stability.

Application of Methods of Decentralized Systems in Management in Lean Manufacturing

Abstract The emergence and subsequent popularization of lean manufacturing have become one of the most significant for improving the efficiency and productivity of operations. The use of lean manufacturing tools and methods leads to the elimination of waste in the organization. Traditional information systems that allow organizations to share information about resources while managing process performance and traceability have a number of disadvantages such as security, interoperability, and transparency. Currently, distributed ledger technology (block-chain) is widely used for this purpose. This article presents a study of decentralized management of the implementation of a distributed ledger infrastructure, which is selected based on the characteristics of the production system. This study proposes a framework that analyzes lean production methods using simulation and data envelopment analysis (DEA) to accommodate the underlying multi-objective decision-making problem. The current study examines the impact of the simultaneous application of RCA technology, lean manufacturing methods, and distributed ledger technology on the total time, costs, and time of production processes.

Solving an energy resource management problem with a novel multi-objective evolutionary reinforcement learning method

Microgrids have become popular candidates for integrating diverse energy sources into the power grid as means of reducing fossil fuel usage. Energy Resource Management (ERM) is a type of Unit Commitment problem, where a player operates a microgrid with diverse renewable generators integrated with an external supplier. Calculating the economic dispatch of each committed unit on a planning horizon is an NP-hard problem, and therefore, finding an exact solution is difficult. This paper presents a multi-objective solution to the ERM problem from the perspective of battery operation and external supplier dispatch. First, a novel multi-objective decision problem modeling is proposed that considers three objectives: cost, greenhouse gas emissions, and battery degradation. This framework involves a learning agent that controls the depth of discharge of a Lithium-Ion battery. To address the proposed problem, a new multi-objective algorithm called Multi-Objective Evolutionary Policy Search (MEPS) is introduced. The proposed algorithm uses NeuroEvolution of Augmenting Topologies structure to evolve artificial neural networks for estimating action-preference values considering multi-objective rewards. The MEPS performance is evaluated on both standard and newly-proposed benchmark problems, using the hypervolume as the evaluation metric. When compared to standard deep reinforcement learning, results showed that MEPS provides cost-effective, environmentally friendly, and efficient energy storage management solutions. Furthermore, MEPS effectively solves the proposed ERM problem by finding neural networks with a small number of nodes and connections, which are suitable for use in embedded control systems. Overall, MEPS proved to be a promising multi-objective approach in the transition to clean energy resources.

Monte Carlo tree search-based deep reinforcement learning for flexible operation & maintenance optimization of a nuclear power plant

Nuclear power plants (NPPs) are required to operate on a flexible profitable production plan while guaranteeing high safety standards. Deep reinforcement learning (DRL) is an effective method to find the most profitable operation & maintenance (O&M) strategy to adopt in a complex system. However, profit-driven only DRL neglects safety-related issues. In this paper, we propose a DRL approach to solve single-objective sequential decision problems (SOSDPs) and multi-objective sequential decision problems (MOSDPs) to find O&M strategies that trade off reliability and profit. The combinatorial problem related with the training of the RL agent to search for the optimal solution is addressed by Monte Carlo tree search (MCTS), whose performance is compared with the traditionally adopted proximal policy optimization (PPO) & imitation learning (IL). A case study is considered for demonstration.

A Decision-Making Method for Design Schemes Based on Intuitionistic Fuzzy Sets and Prospect Theory

Conceptual design is a key link in the process of complex product design, and it is very important to select the appropriate design scheme; however, there are many types and inaccuracies of the evaluation data, and there is a problem of mutual influence between the evaluation criteria, which leads to unreliable decision making of the optimal solution. In order to solve this problem, a decision-making method based on intuitionistic fuzzy sets (IFS) and prospect theory is proposed. This method can be used for symmetric and asymmetric evaluation data. The evaluation data are classified according to different expression types and unified using intuitionistic fuzzy numbers. The intuitionistic fuzzy prospect value of decision information is calculated using prospect theory, and the prospect transformation of decision information is completed. At the same time, the Gray Relational Analysis (GRA) method and the Criteria Importance Though Intercriteria Correlation (CRITIC) method are used to calculate the subjective and objective weights of the technical and economic evaluation indexes of the product, and the combination weights are given; then, based on the evidence theory, the basic probability distribution of the evidence chain of all conceptual design schemes is synthesized, and the comprehensive prospect evaluation results of the schemes are obtained to complete the optimization of the conceptual design schemes. Finally, the effectiveness of the proposed method is verified by the conceptual design of the chip removal system of the deep hole machining machine tool. This work provides a promising method for decision makers to optimize the design scheme and provides insights into multi-objective decision-making problems.

Phase-wise evaluation and optimization of non-pharmaceutical interventions to contain the COVID-19 pandemic in the U.S.

Given that the effectiveness of COVID-19 vaccines and other therapies is greatly limited by the continuously emerging variants, non-pharmaceutical interventions have been adopted as primary control strategies in the global fight against the COVID-19 pandemic. However, implementing strict interventions over extended periods of time is inevitably hurting the economy. Many countries are faced with the dilemma of how to take appropriate policy actions for socio-economic recovery while curbing the further spread of COVID-19. With an aim to solve this multi-objective decision-making problem, we investigate the underlying temporal dynamics and associations between policies, mobility patterns, and virus transmission through vector autoregressive models and the Toda-Yamamoto Granger causality test. Our findings reveal the presence of temporal lagged effects and Granger causality relationships among various transmission and human mobility variables. We further assess the effectiveness of existing COVID-19 control measures and explore potential optimal strategies that strike a balance between public health and socio-economic recovery for individual states in the U.S. by employing the Pareto optimality and genetic algorithms. The results highlight the joint power of the state of emergency declaration, wearing face masks, and the closure of bars, and emphasize the necessity of pursuing tailor-made strategies for different states and phases of epidemiological transmission. Our framework enables policymakers to create more refined designs of COVID-19 strategies and can be extended to other countries regarding best practices in pandemic response.

Fuzzy decision-making approach of hobbing tool and cutting parameters

Machining performance, such as carbon emission, machining time, and quality, can be improved by selecting appropriate tools and cutting parameters for gear hobbing. However, the importance machining performance is usually difficult for users to quantify. To resolve the decision-making problem of hobbing tools and cutting parameters in fuzzy evaluation, a multi-objective fuzzy decision-making approach for hobbing parameters is proposed. It uses spectral clustering, the multi-objective weighted mean of vectors (MOINFO) algorithm and the fuzzy technique for ordering preference by similarity to the ideal solution (TOPSIS) algorithm. First, spectral clustering is used to confirm the initial range of the hobbing parameters based on past machining cases. MOINFO is then applied to search for nondominated hobbing parameters based on the optimization model, taking the carbon emission, cutting time, and quality as the objectives. Finally, the nondominated hobbing parameters are sorted by fuzzy TOPSIS according to a user evaluation, and the hobbing parameter ranking is used to guide hob selection and machining. Decision-making experiments on the hobbing parameters are performed with actual hobbing. The comprehensive evaluation index CEI result for the proposed approach is 0.19. Compared with other methods, the proposed approach exhibits certain advantages. These results demonstrate that the multi-objective decision-making problem of hobbing parameters under fuzzy evaluation can be solved using the proposed approach.

A new multi-objective decision-making method with diversified weights and Pythagorean fuzzy rough sets

In reality, there are many multi-objective decision-making problems with fuzziness and uncertainty, such as the choice of clinical treatment schemes. To effectively cope with such problems, it is urgent to scientifically handle fuzzy and uncertain information. Given the fact that the Pythagorean fuzzy set is a useful tool to tackle fuzzy information, and the rough set theory is superior in dealing with uncertain data, this paper aims to establish a novel decision-making method for scheme selection from the perspective of Pythagorean fuzzy rough sets. Pythagorean fuzzy numbers (PFNs) are first employed to express fuzzy preference information, and then multi-objective Pythagorean fuzzy information systems are introduced. Subsequently, we develop a new fuzzy rough set using an improved correlation coefficient method to deal with the uncertainty of PFNs. Furthermore, considering the diversified weights, we present a weight calculation method based on approximate accuracies. Meanwhile, a weighted fuzzy rough set model is constructed to achieve multi-objective information fusion. Accordingly, a multi-objective decision-making method is put forward with the help of overlap functions. Finally, we apply the built method to three cases from different decision scenarios to illustrate the applicability of the method. By some comparative analyses, the effectiveness and superiority of the method are also exhibited. Moreover, we discuss the establishment of decision support systems regarding scheme selection. On the whole, the work of this paper is helpful to support the solution of complex multi-objective decision-making problems.

Development of AHP-embedded Deng’s hybrid MCDM model in micro-EDM using carbon-coated electrode

Abstract The use of thin film-coated micro-tool electrode introduced a new technical solution to bring the best technical–economic efficiency in the field of non-conventional machining processes. Multi-objective optimization of process parameters in micro-electrical discharge machining (EDM) is very limited with the help of thin film-coated micro-tool, which brought benefits for the industry. In this research work, the technological parameters such as voltage (V), spindle rotation, and capacitance (C) in micro-EDM using carbon-coated electrode were optimized, and Z-coordinate depth of cut and tool wear rate were selected as response variables. Taguchi is combined with Deng’s method to solve this multi-objective decision problem. The results showed that Taguchi–Deng’s method is the right solution for multi-target decision in micro-EDM using the coated electrode. C is the strongest influence on optimal efficiency, and it is the smallest with V. The machining quality under optimal conditions is improved.

A novel multilevel decision-making evaluation approach for the renewable energy heating systems: A case study in China

Renewable energy (RE) sources are important alternatives to mitigate the energy crisis and achieve sustainable development. Appropriate selection of RE system solutions is extremely crucial. Selection of the best RE technology requires the consideration of conflicting qualitative and quantitative evaluation criteria. Many evaluation criteria are judged with subjectivity and uncertainty. However, the description of uncertainty in the evaluation process remains a large research gap. Therefore, a novel combined evaluation method is developed to describe and visualize the uncertainty in the assessment process. The proposed evaluation method is tested for RE heating system selection. The RE systems are evaluated based on five dimensions and 15 evaluation indicators. This multidimensional indicator framework not only includes the three basic evaluation groups of energy, economy, and environment, but also extends to the performance of technology and policy. The combined weights of the evaluation indicators consist of objective weights and subjective weights. The objective weights are obtained by the Criteria Importance Through Intercriteria Correlation (CRITIC) method and subjective weights are calculated by the improved Fuzzy Analytic Hierarchy Process (FAHP). The set pair analysis (SPA) is introduced to assess the performance of different RE systems. It considers the uncertainty of indicator performance. A novel approach to visualizing the fuzziness of SPA evaluation is developed using the cloud model. Finally, the RE system ranking calculated by the proposed method is performed. The originality of this work is offering a promising method for RE selection and clarifying the degree of ambiguity in the evaluation process. It helps decision makers have an exact idea about the accuracy of the evaluation. It provides insights into multi-objective decision-making problems.

Multi-objective Reinforcement Learning – Concept, Approaches and Applications

Real-world decision-making tasks are generally complicated and require trade-offs between multiple, even conflicting, objectives. As the advent and great development of advanced information technology, it has evolved into using reinforcement learning (RL) algorithms to tackle the multi-objective decision making (MODM) problems. In this paper, we will first identify the basic concepts and factors when modelling the MODM tasks with reinforcement learning, and then review the traditional RL, such as Sarsa, Q-Learning, Policy Gradients, Actor-Critic, Monte-Carlo learning, and modern deep RL algorithms applied in this process. Furthermore, the specific practical scenarios described in MODM problems will be summarized through analyzing some typical articles. Finally, the future trends of multi-objective reinforcement learning will be discussed.

Improved multi-objective differential evolution algorithm based on a decomposition strategy for multi-objective optimization problems

Many real-world engineering problems need to balance different objectives and can be formatted as multi-objective optimization problem. An effective multi-objective algorithm can achieve a set of optimal solutions that can make a tradeoff between different objectives, which is valuable to further explore and design. In this paper, an improved multi-objective differential evolution algorithm (MOEA/D/DEM) based on a decomposition strategy is proposed to improve the performance of differential evolution algorithm for practical multi-objective nutrition decision problems. Firstly, considering the neighborhood characteristic, a neighbor intimacy factor is designed in the search process for enhancing the diversity of the population, then a new Gaussian mutation strategy with variable step size is proposed to reduce the probability of escaping local optimum area and improve the local search ability. Finally, the proposed algorithm is tested by classic test problems (DTLZ1-7 and WFG1-9) and applied to the multi-objective nutrition decision problems, compared to the other reported multi-objective algorithms, the proposed algorithm has a better search capability and obtained competitive results.

A selection scheme for energy storage to participate in primary frequency regulation of new energy

The choice of energy storage is the key content of the energy storage plan and design to participate in the primary frequency regulation of new energy. It involves many indicators such as safety, primary frequency modulation adaptability, economy, and environmental performance, and is a complex multi-objective decision-making problem. This paper analyses the characteristics of each index and constructs a two-level decision-making system. According to expert scores and practical application requirements, the weight of each decision-making index is obtained by the AHP method. The TOPSIS method is further used to select energy storage according to the weight of each index and related parameters.

A new interactive fuzzy programming with priority control for multi-objective decision-making problems

Fuzzy multiple objective linear programming (FMOLP) approaches have been used to be solved many applications of multi-objective decision-making (MODM) problems. Several methods have been proposed including max-min approaches, preemptive approaches, and weighted approaches. However, they have some limitations in use; some may not be able to obtain efficient solutions or can give only a few solutions. Some methods need to be solved in two steps or specify the target level, which may be difficult for decision-makers (DMs). In this research, a new single-phase interactive fuzzy programming approach with priority control that can find several efficient solutions is proposed. It is not necessary to specify the target value for each objective and it can solve with only one step for each solution. The DM can easily select the appropriate solutions from a set of efficient solutions. This method is different from existing single-phase approaches by controlling the satisfaction level of the last priority instead of using weight additive. Simple examples and a practical example of a perishable product supply chain network were tested to show the effectiveness of the proposed model. The performance of the proposed method was also compared with existing methods to verify and validate the model.

A New Decision-Making GMDH Neural Network: Effective for Limited and Fuzzy Data.

This paper presents a new approach to solve multi-objective decision-making (DM) problems based on neural networks (NN). The utility evaluation function is estimated using the proposed group method of data handling (GMDH) NN. A series of training data is obtained based on a limited number of initial solutions to train the NN. The NN parameters are adjusted based on the error propagation training method and unscented Kalman filter (UKF). The designed DM is used in solving the practical problem, showing that the proposed method is very effective and gives favorable results, under limited fuzzy data. Also, the results of the proposed method are compared with some similar methods.

Supporting strategy selection in multiobjective decision problems under uncertainty and hidden requirements

Decision-makers are often faced with multi-faceted problems that require making trade-offs between multiple, conflicting objectives under various uncertainties. The task is even more difficult when considering dynamic, non-linear processes and when the decisions themselves are complex, for instance in the case of selecting trajectories for multiple decision variables. These types of problems are often solved using multiobjective optimization (MOO). A typical problem in MOO is that the number of Pareto optimal solutions can be very large, whereby the selection process of a single preferred solution is cumbersome. Moreover, preference between model-based solutions may not be determined only by their objective function values, but also in terms of how robust and implementable these solutions are. In this paper, we develop a methodological framework to support the identification of a small but diverse set of robust Pareto optimal solutions. In particular, we eliminate non-robust solutions from the Pareto front and cluster the remaining solutions based on their similarity in the decision variable space. This enables a manageable visual inspection of the remaining solutions to compare them in terms of practical implementability. We illustrate the framework and its benefits by means of an epidemic control problem that minimizes deaths and economic impacts, and a screening program for colorectal cancer that minimizes cancer prevalence and costs. These examples highlight the general applicability of the framework for disparate types of decision problems and process models.

Comparing chair time required for systemic agent drug reaction (DR) management to allocated chair time for systemic protocol.

40 Background: Chemotherapy scheduling and determining associated nursing resources are complex multiobjective decision problems. Management of DRs are thought to take additional chair and nurse time and should be accounted for in appointment scheduling in order to having an efficiently running clinic. BC Cancer – Victoria is a tertiary care regional site of BC Cancer in Victoria, British Columbia, Canada and currently schedules patients using a regimen-based resource intensity model (Greene et al, 2012). Methods: DRs were tracked in real time over 3 months, where BC Cancer ID, reaction, protocol, agent, cycle number and time of reaction were recorded. Our electronic medical record was interrogated to determine start and end of chair time. We then compared the actual chair time required to manage patients with a DR to scheduled chair time. Results: Thirty reactions occurred, including DRs to paclitaxel (43%), rituximab (13%), docetaxel (10%), nivolumab (7%), oxaliplatin (7%), pertuzumab (7%), bendmustine (3%), and daratumumab (3%). The reactions most commonly occurred with cycles 1 or 2. The averages for total chair time during DR, time from arrival to reaction, and time from reaction to discharge were 281 min (+79.5), 106 min (+ 55), and 171 min (+ 94), respectively. An average of 449 min/month extra were required for DR management above scheduled chair time. Conclusions: The agents precipitating DRs were common and predictable. An average of 7 hours and 29 minutes per month above scheduled chair time was required to manage DRs. This information could be used to more efficiently schedule chair and nursing time. This also highlights the need to minimize the number of DRs as management is resource intensive.[Table: see text]

Learning-based multi-objective evolutionary algorithm for batching decision problem

This study investigates a multi-objective batching decision problem that arises in batch annealing operations in the iron and steel industry. The problem concerns selecting coils from a set of waiting coils to be annealed to form batches so as to maximize two conflicting objectives: product quality and equipment utilization. In this study, a multi-objective optimization model is formulated, and a data-driven method is developed to determine the technological parameters for the model. To improve the efficiency and effectiveness of the solution process, we propose a learning-based multi-objective evolutionary algorithm (LBMOEA) with novel evolution operators and a learning-based solution space reduction strategy. To more quickly solve the problem, a clustering method is adopted to achieve a parallel mechanism in the LBMOEA. In computational experiments on 20 randomly generated instances, the results demonstrate that the above evolution operators and strategy are effective, and that the clustering method can reduce the average time cost by 58.76%. For 15 practical production instances, the results illustrate that the LBMOEA with the clustering method is superior to the non-dominated sorting genetic algorithm II (NSGA-II) and multi-objective evolutionary algorithm based on decomposition (MOEA/D), and shows good potential for application in practical production.

A subjective-objective integrated multi-objective decision-making method for reservoir operation featuring trade-offs among non-inferior solutions themselves

The key to formulating a multi-objective reservoir operation scheme is coordinating the high/low priorities and achieving balance among various targets accordingly in the decision-making process. However, the traditional decision-making methods are either completely subjective or neglect the decision-maker's preferences, making it essential to optimize these methods to address such unsatisfactory aspects. Therefore, through drawing on some universally acknowledged economic theories and studying the geometric relationships among the Pareto solutions to such multi-objective decision-making problems, the concepts of “replacement rate”, “profit-loss sensitivity ratio”, and “preference equilibrium degree” are introduced, deduced, and elaborated on in this paper. On the basis of them, the preference decision method (PDM), entrusted with strict mathematical and physical connotations, is constructed and derived in detail, especially for addressing the bi-objective and tri-objective scenarios. The PDM can quantify the complex internal feedback relationship among various targets, providing objective support for decision-makers in the subjective trade-offs of the interests of multiple subjects and achieving the unity of subjective and objective. Furthermore, the proposed method is applied to the Wujiang cascade reservoirs in China, through which the benefits of power generation, water supply and ecological protection are enabled to be considered simultaneously. The application shows that the PDM can effectively shrink the Pareto set, which greatly reduces the difficulty of decision-making, and the obtained results prove feasible and satisfactory. In addition, the PDM is based entirely on the Pareto set itself, making it simple to utilize, and can be extended into solving other multi-objective decision-making problems.

Genetic algorithm optimised Hadamard product method for inconsistency judgement matrix adjustment in AHP and automatic analysis system development

The analytic hierarchy process (AHP) is an important method to solve the multi-objective decision-making weight problem. However, due to the subjective judgement and selection preference of decision-makers, the consistency of the judgement matrix is inevitably poor. Hence, to improve the reliability of decision-making results, it is necessary to adjust the consistency of the judgement matrix. We propose a genetic algorithm optimised Hadamard product (GAOHP) to judge the consistency of the matrix. This method converts the original matrix consistency adjustment problem into an optimisation solution problem and uses a meta-heuristic algorithm to search the global optimisation quickly. Our method, which significantly improves the computing efficiency and realises that the judgement matrix after adjustment satisfies the basic consistency, has fully retained the judgement intention of the decision-makers. It offers the dual advantages of preserving the original intention of decision-makers and greater efficiency of operation. Finally, we developed an automatic analysis system based on MATLAB app designer to realise the rapid adjustment of the consistency of judgement matrix, which is conducive to providing an analysis system with simple and stable operation for AHP analysis.