MISANU

Seminar for
DECISION MAKING – THEORY, TECHNOLOGY AND PRACTICE

PROGRAM

Plan rada Seminara Odlučivanje - teorija, tehnologija, praksa za JUN 2021.

ČETVRTAK, 03.06.2021. u 13:00, Live stream Niš
Dragan Pamučar, University of Defence, Military academy in Belgrade
NEWLY DEVELOPED MULTI-CRITERIA FRAMEWORKS: LBWA METHOD FOR DEFINING WEIGHTS AND MARCOS METHOD FOR EVALUATION
The lecture on the topic of newly developed multi-criteria frameworks will be composed of three parts. In the first two parts, new multi-criteria frameworks for robust decision making will be presented, while in the third part, the methodology for validation of the obtained results and sensitivity analysis of the multi-criteria models to variation of input parameters will be presented.
To objectify the decision-making process and improve the performance of existing models, scientists are actively exploring the application of new multi-criteria methodologies. One of the more recent methodologies for determining the weight coefficients of the criteria is the Level Based Weight Assessment (LBWA) model (Žižović & Pamčar, 2019). The LBWA method enables experts from different fields to define the relationship between the criteria and rational decision-making. The technique can be applied in practical cases in specialized decision-making support systems and alternative dispute resolutions in virtual environments. The LBWA model has several key advantages over other subjective models based on pairwise comparison of criteria, which include the following: (1) the LBWA model allows the calculation of weight coefficients with a small number of criteria comparisons, only n-1 comparison; (2) The algorithm of the LBWA model does not become more complex with the increase of the number of criteria, which makes it suitable for use in complex multi-criteria (MCDM) models with a large number of criteria; (3) By applying the LBWA model, optimal values of weight coefficients are obtained with simple mathematical apparatus that eliminates inconsistencies in expert preferences, which are tolerated in specific subjective models (Best Worst Method - BWM and Analytic Hierarchy Process - AHP); (4) The elasticity coefficient of the LBWA model enables, after comparing the criteria, additional corrections of the values of the weight coefficients depending on the preferences of the decision-makers. This feature of the LBWA model enables sensitivity analysis of the MCDM model by analyzing the effects of variations in the weights of the criteria on the final decision.
In addition to the LBWA method, which is used to determine weighting coefficients, one of the newer methodologies for evaluating solutions is the Measurement Alternatives and Ranking according to the Compromise Solution (MARCOS) method (Stević, Pamučar, et al., 2020). In addition to the fact that the MARCOS method is more recent, due to its advantages, it has taken a significant place among the increasingly applied multi-criteria techniques due to the robustness of the obtained results in a dynamic environment. The MARCOS method integrates three starting points for providing a robust decision: (1) Defining reference points (ideal and anti-ideal values); (2) Determining the relationship between alternatives and ideal/anti-ideal values; (3) Defining the degree of usefulness of alternatives concerning the ideal and anti-ideal solution.
Validation of the results of multi-criteria models represents the final phase before the implementation of the final decision. Therefore, it is necessary to examine the proposed solution's quality and single out the dominant alternative from the considered set. One of the methodologies for checking the robustness of the solution of multi-criteria models will be presented in this lecture, which includes the following phases: (1) The influence of the change in the weight of the criteria on the ranking results; (2) Simulation of the dynamic environment through a change of input parameters of initiation matrix; and (3) The impact of changes in the parameters that are defined in the multi-criteria model based on the subjective assessments of the decision-maker.

ČETVRTAK, 10.06.2021. u 13:00, Live stream Niš
Ivana Veličkovska, Mathematical Institute of the Serbian Academy of Science and Arts
SELECTION OF STRATEGIES FOR ECONOMIC DEVELOPMENT IN COAL MINING USING A HYBRID SWOT-AHP METHOD
Competitiveness on the global market is reflected in the struggle for natural resources, amongst the most important ones are coal, oil and natural gas. However, coal is a conventional energy source and as such is non-renewable and it is necessary to devise a sustainable plan for its use so that future generations can exploit this resource. Serbia is a country rich in coal reserves and accounts for majority of energy, thereby coal supply is a critical factor for future functioning of the energy system in Serbia. Although there are strong tendencies towards the use of renewable energy sources, coal remains the most important source for electricity generation in Serbia and beyond. Having in mind numerous issues that are reported in the national energy sector, the aim of this paper is to propose and prioritize future strategies for economic development of a public enterprise for underground coal exploitation located in Serbia using multicriteria decision-making methodology. Considering complex business environment involving a number of influential factors, both internal and external, are taken into account when creating a company's strategy. For defining influencing factors, a very useful tool called SWOT analysis is used. Four criteria and nineteen sub-criteria are identified. By considering the criteria and sub-criteria of SWOT analysis, six business strategies are proposed. Using the AHP method of multi-criteria decision-making, prioritization is carried out and the obtained results show that the most important proposed strategy that has priority in implementation is the strategy of diversification.

ČETVRTAK, 17.06.2021. u 13:00, Live stream Niš
Dušan Tatić, Mathematical Institute of the Serbian Academy of Science and Arts
AN AHP BASED MODEL FOR EVALUATION OF DEVICES FOR APPLICATION OF AUGMENTED REALITY IN INDUSTRY
In Industry 4.0, advanced technologies such as augmented reality (AR) are well accepted and efficiently applied in different industrial environments for various purposes. The related hardware solutions for their implementation are different and depend on the type of AR applications. Consequently, choosing the best suited and most effective AR device for a particular AR application in a concrete industrial process is a very complex task. A multi-criteria selecting model based on Analytic Hierarchy Process (AHP) method could be a useful tool for solving this task. We present and discuss such a model proposed in the paper Elia, V., Gnoni M. G., Lanzilotto, A., Evaluating the application of augmented reality devices in manufacturing from a process point of view: An AHP based model, Expert Systems with Applications 63 (2016), 187-197.

ČETVRTAK, 24.06.2021. u 13:00, Live stream Niš
Dragana Radojičić, Mathematical Institute of the Serbian Academy of Science and Arts
A MULTICRITERIA OPTIMIZATION APPROACH FOR SELECTION OF THE NEURAL NETWORK-BASED LIMIT ORDER BOOK MODEL
In this research, we study the in formativeness of features extracted from a limit order book data to label market data vector by using different neural networks. More precisely, the idea is to use a neural network to classify each vector of the order book data into one of the labels from the set S={buy, sell, idle}. Furthermore, we cross-compare the performances of the models based on different networks, namely the Gated recurrent unit (GRU), Long short-term memory (LSTM), and Recurrent neural network (RNN) network via the multicriteria decision-making method.

Predavanja su namenjena sirokom krugu slusalaca, ukljucujuci studente redovnih i doktorskih studija. Seminar će se održavati svakog drugog četvrtka od 13:00 - 14:00^h, CIITLAB, Elektronski fakultet Niš, Aleksandra Medvedeva 14, Niš

dr Lazar Velimirović
Rukovodilac seminara

dr Petar Vranić
Sekretar seminara