Comparison of linear and vector data normalization techniques in decision making for learning quota assistance

Data normalization is essential for all kinds of decision-making problems, and a lot of effort has been spent on the development of normalization models in multi-criteria decision making (MCDM), but despite all this, there is no definite answer to the question: Which is the most appropriate technique?. This paper compares the popular normalization techniques: Linear Normalization (LN) and Vector Normalization (VN) using VIšekriterijumsko KOmpromisno Rangiranje (VIKOR) Method. The beneficiaries dataset of learning quota was collected of 399 students sample through observation (drive-test measurements and online questionnaires) to obtain information on criteria data including attributes in online learning during the Covid-19 pandemic. The ranking results for vector vs linear normalization show how ranking is affected. The difference in the selection of the best alternative (rank) shows that there are differences in vector and linear assessments that are influenced by the max-min criterion value which has an impact on the ranksum results (benefit/cost). This test clearly shows how important it is to use an appropriate (normalized) representation of the model because there will often be a criterion where "the higher the better" while for others (cost) "the lower the better".