Forecasting Cattle Population: A Case Study of Türkiye

Cattle breeding is of critical importance in meeting the animal protein needs of the increasing population due to its significant contribution to meat and milk production, which are the main animal protein sources. In addition, cattle breeding has important potential for both the agricultural economy and the general economy in terms of the production and export of value-added agricultural goods and processed products, especially for countries with a large number of cattle. In order to maximize these and similar benefits, to evaluate the structural problems in the livestock sector and to implement effective policies to increase the cattle population to optimum levels, it is of great importance to make data-based decisions and therefore produce sufficient and necessary data. Achieving this will be possible not only with existing data but also by making forward projections with solid scientific methods and estimating the necessary data to plan for the future now. The purpose of this research is to estimate the number of cattle for the next ten years by comparing the results of the artificial neural networks (ANN) and autoregressive integrated moving average (ARIMA) models, using Türkiye's cattle number at the beginning of the year for the years 1930-2024. According to the research results, ARIMA had a greater ability to forecast than ANN. Box-Jenkins method was used in the ARIMA estimations. The ARIMA (1,1,0) model was determined to be the most appropriate model for the data, and it was estimated that the number of cattle at the beginning of the year will increase in the next ten years, reaching 17313762 head in 2025 and 17317161 head in 2033, representing a 5.5% increase in the ten-year period.