Demand Forecasting Software for Medical Products

Abstract

this study evaluates the sufficiency of available data for factor analysis to develop an equation for calculating planned annual requirements, aiming to reduce labor costs. The implementation uses C# and Microsoft Excel. The dataset includes such variables as the number of patients with diabetes mellitus in the region, the average annual change in their numbers, the number of drugs received, the gender and age of the patients, and the distribution of patients into three categories of government benefits they are entitled to. Simplifications are applied where they do not significantly impact the results. The findings highlight the potential of this approach. We are going to develop and train a neural network to obtain more precise equations. The study also emphasizes the need to explore automated model creation for factor analysis to assess the current model’s quality. For cases where sufficient data is available to train a neural network that reliably predicts planned requirements, adopting neural network solutions is recommended over the current approach.