Türkiye’deki Tiftik Üretimi Değişiminin Regresyon Analizi ile İncelenmesi

Abstract

The aim of this study is to investigate the changes in the number of Angora goat and the amount of mohair obtained from Angora goats, which is one of the Turkey's important gene sources from Central Asia to present day, between 1991-2019 in Turkey via different regression models and to evaluate the results. For this purpose, simple linear, quadratic, cubic, inverse, and logarithmic regression models are used in the study. In order to compare the regression models and determine the most suitable model, the square root of the mean squares (Root Mean Square error-RMSE), the determination coefficient (R2) and the adjusted determination coefficient (AdjR2) were used as comparison criteria. Accordingly, R2 values obtained from simple linear, quadratic, cubic, inverse and logarithmic regression models for the number of Ankara goats shorn are 0.628, 0.99, 0.99, 0.74, 0.90, and RMSE are 135288.27, 25651.46, 18966.20, 114681.75, 71592.54, respectively. For Mohair production, R2 values are 0.61, 0.98, 0.99, 0.75, 0.88, while RMSE is 208.99, 41.84, 32.64, 167.85 and 114.32 respectively. In the number of angora goats, R2 values are 0.70, 0.99, 0.99, 0.73, 0.93, while RMSE is 165264.22, 32818.49, 23410.64, 155421.63 and 79544.79, respectively. In parameter estimates, the most appropriate model according to the highest R2 value and the lowest RMSE value is the cubic regression model. According to the cubic regression model, the estimated number of Angora goats that will be shorn, the number of Angora (mohair) goat, and mohair production in Turkey will be 254307 and 275431 heads, 268321 and 287846 heads, and 439 and 474 tons in 2020 and 2021, respectively. Although in the coming years, an increase in mohair production is projected in Turkey, this is much lower than expected.