Modelling seasonal allergy incidence based on pollen monitoring data and its application in pharmaceutical logistics

The article aims to develop and test a model for forecasting seasonal allergic diseases based on aeropalynological monitoring data in Perm. The methods include mathematical modelling, machine learning techniques (LSTM, XGBoost, regression models) and time series analysis. The evidence base covers long-term data on pollen concentrations of major allergenic plants, meteorological parameters and statistics on the supply of antihistamines to medical and pharmaceutical organisations in Perm Krai. The models built have shown high forecast accuracy, which has made it possible to formulate practical recommendations for inventory management in retail pharmacies. The results obtained are useful for optimising pharmaceutical logistics, improving healthcare system preparedness, and minimising the consequences of seasonal allergy exacerbations, ensuring more efficient resource planning in the region.