Summary and Critical Analysis of Multi-objective Decision Making - Term Paper Example

Summary and Critical Analysis of Multi-objective Decision Making Introduction Recent years have seen the increasing applicability of multi-objective analysis in solving a number of economic, managerial, and construction problems (Brauers, Zavadskas and Peldschus). Real world problems are constrained by a diverse amount of structures and processes, rarely appropriate variable representations, and conflicting development objectives. Decisions are made by different stakeholders with a variety of interests and values, making the decision-making process even more complicated. Multi-objective decision making (MODM) grants provisions to evaluate different alternatives based on a set of objectives that will facilitate the choosing of the best one (Lu). Multi-objective decision making problems require a set of criteria prior to implementation. First, objective functions must be normalized. That is, objective functions must have the same physical meaning to all the stakeholders involved, which likewise means that units and scales used in the measurement of these function must be consistent (Pedrycz, Ekel and Parreiras). Next, the appropriate principle of optimality must be chosen. The principle of optimality sets the properties of the optimal solution and answers in such a ways that the optimal solution bests all other possible solutions (Pedrycz, Ekel and Parreiras). Third, the priorities of objective functions have to be considered. Different objective functions have different importance and thus, higher priority must be given to more important objective functions (Lu). Based on these considerations, this paper will provide a critical analysis of a MODM model which covers economics and managerial applications. A summary of the model will be presented, and its strengths and weaknesses will be highlighted. The final portion of this paper will present an outline of a possible research project that will involve MODM. MODM in Supplier Selection and Order Allocation Supplier selection is an area that utilizes multi-objective decision making because it involves a number of criteria that must be considered, such as capacity constraints. When a business is presented with a choice from among different suppliers, it must decide on which supplier is the best one and on how much should be purchased from each of the selected suppliers. Demirtas and Ustun (2008) integrated analytic network process approach and multi-objective mixed integer linear programming (MOMILP) to take into consideration the physical and intangible factors when choosing the best suppliers and determining the optimum quantities that must be purchased from each in order to maximize the total purchase value while, at the same time, minimize the costs incurred. The resulting objective functions and constraints of the model are given below: subject to: where, Wi Normal weights of the ith supplier Xi Order quantity for the ith supplier Vi Capacity of the ith supplier Yi Binary integer (0—if the order is given to the ith supplier, 1—if the order is not given to the ith supplier Ci Unit cost for the ith supplier Oi Order cost for the ith supplier D Demand for the period qi Defect rate of the ith supplier It should be noted that this model was compared to a couple more models in the research paper but these are not part of the actual coverage of the critical analysis and were thus not included in the discussion anymore. Insights from two other models (Brauers, Zavadskas and Peldschus; Higgins, Hajkowicz and Bui), were incorporated within the discussion. Strengths of the model This MODM model provides managers the means to scientifically make crucial decisions in different business set ups (Pedrycz, Ekel and Parreiras). One is assured of the reliability of this model as data was based on actual figures, which were then used to improve the consistency of the system (Demirtas and Ustun). These figures are actually used for the quantitative data in comparison. The model shows that a certain level of care was used to approximate the representation of real life concepts and variables. Another good thing about the model is the minimal number of objective and constraints function. This would mean that up to a certain degree, this model is quite manageable and is able to provide an option for managers to scientifically support certain managerial decisions. In addition, the model provided sufficient tools to catch the effects of the same level criteria on themselves and the effects of the alternatives on the criteria. On the significance of the actual use of the model, one should recall that one of the most important activities with supply chain partners in a business-to business (B2B) network is purchasing (Pedrycz, Ekel and Parreiras). In fact, it is also the main point of contact with supply-chain partners. Thus, the authors of the model were right to come up with a model that will be able to simplify and optimize the outcome of this activity. The decisions made at the purchasing stage of the B2B transaction create a major impact with the rest of the supply-chain system. Weaknesses of the model As with most MODM models, the benefits of such a model can be fully taken advantafe of by experts with sophisticated analytical skills in the particular field (in this case, economics and management). The model practically requires that the user is able to identify the existing (and appropriately applicable) problems and match these together. The model is highly mathematical in nature that average-level managers may not be able to catch up with the algorithms involved in the model. Another weakness of the model is that it failed to incorporate a number of important considerations in the choice of supplier. Some of these considerations are: management capability, production flexibility, design and technological capability, financial stability, experience and geographical location. These considerations create quite an impact to the choice of supplier because they take into consideration the integration capabilities of prospective suppliers, thus providing better qualifications for choice of supplier especially in a scenario of integrated supply chain management (Lu). Multi-objective decision models must represent a real-world scenario as much as possible and as such, the aforementioned considerations should be integrated into the model in order to improve its capability to provide a most relevant and optimal solution in the choice of supplier. Research Project Outline: Multi-objective Decision Making in Selecting Projects in Research and Technology I. Introduction II. Review of Relevant Literature regarding Multi-objective Decision Making III. Building of the Model A. Objective functions B. Constraints C. Notations D. Possible alternatives IV. Provision for a Numerical Example V. Conclusions and Recommendations VI. References References Brauers, Willem, et al. "Multi-objective decision making for road design." Transport 23 (2008): 183-193. Demirtas, Ezgi and Ozden Ustun. "An integrated multiobjective decision making process for supplier selection and order allocation." Omega: The International Journal of Management Science 36 (2008): 76-90. Higgins, Andrew, Stefan Hajkowicz and Elisabeth Bui. "A multi-objective model for environmental investment decision making." Computers and Operations Research 35 (2008): 253-266. Lu, Jie. Multi-objective group decision making: Methods, software, and applications with fuzzy set techniques. Imperial College Press, 2007. Pedrycz, Witold, Petr Ekel and Roberta Parreiras. Fuzzy multicriteria decision-making: Models, methods and applications. Hoboken, NJ: John Wiley and Sons, 2011. Read More