ABSTRACT
To handle future uncertainty, it is critical to utilise time series forecasts while making decisions. Efficient forecasting is seen as a significant prerequisite for successful management and organization in a variety of social, knowledge, human and natural sciences and related fields of application. Various forecasting approaches have been proposed to deal with the increasing uncertainty and complexities associated with domain-specific forecasting problems. When deciding on a forecasting algorithm, decision-makers must consider various aspects of the prediction process, such as the duration of the forecast horizon, the aim of forecasting, the frequency, structure and nature of the data. The focus of this paper is to implement three MADM techniques namely, Analytic Hierarchy Process (AHP), Technique for Order of Prefer-ence by Similarity to Ideal Solution (TOPSIS) and VlseKriterijumska Opti-mizacija I Kompromisno Resenje (VIKOR) on a practical application using Python to propose an MCDM approach to evaluate and rank quantitative demand forecasting models. The rankings are based on error measurements like Mean Squared Error (MSE), Root Mean Squared Error (RMSE), etc. which play an instrumental role in forecasting algorithms. The conclusion outlines the best time-series method that complies with the given constraints.
