Application of Method for Non-Linear Scaling of Multi-Criteria Decision Making Attribute Values

When analyzing Multi-Criteria Decision Making (MCDM) problems, many methods only consider linear attribute values and scale them accordingly. Often there are attributes that do not follow a linear translation and therefore should be treated differently. Non-linear scaling is difficult to implement and even harder to interpret, but necessary in order to represent the data in a realistic way. This study compares the proposed non-linear and the standard linear scaling methods using TOPSIS as the MCDM technique of choice. A baseline TOPSIS is performed using linear scaling and then using the non-linear method proposed here. The results show a change in the outcome of the analysis which aligns with the expected results given the group solicited scaling values collected.