Answers are grounded in this paper and cite their sources.
This study examines the effectiveness of two statistical methods - the multiplicative Holt-Winters (HW) model and the autoregressive (AR) model - in forecasting the costs of economic tasks carried out in the forest districts of the Regional Directorate of State Forests in Piła. Historical data on activities such as artificial regeneration, crop care, pre-cutting and land reclamation were used, analyzing their seasonality and temporal dynamics. A comparison was made between the accuracy of the two models' predictions using MAE and MAPE error measures. The results show that the HW model performs better for costs showing regular seasonality, while the AR model proves more effective in forecasting variables of a more random nature. The study confirms the applicability of classical time series forecasting methods as a tool to support the process of financial planning in forest management, with the need for further optimization and personalization of models for individual units.
No takes yet. Share an insight, caveat, or question.
Witulska et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: