2018, Cilt 34, Sayı 3, Sayfa(lar) 142-149
Forecasting of commercial egg production in Turkey with Box-Jenkins and Winter’s Exponential Smoothing Methods
Özlem Kaymaz
Ankara Üniversitesi, İstatistik Bölümü, Ankara, Türkiye
Keywords: Box-Jenkins Methods, forecast, SARIMA, time series, Winter’s Exponential Smoothing Method

Aim: The developments in the poultry industry, due to its closeness to an important part of Turkey's potential export markets has increased the interest in this sector. The aim of this study was to evaluate between January (2010) and February (2018) of time series related to commercial egg monthly production in Turkey, with Box-Jenkins and Winter's exponential smoothing method and comparison of the forecast models.

Materials and Methods: Data was obtained from Turkey Statistical Institute (TUIK) database which related to commercial egg monthly production in Turkey between 2010 to 2018. In this study, the predictions obtained by using the seasonal autoregressive integrated moving average (SARIMA) and Winter's exponential smoothing method and the forecast models were compared with RMSE, MAPE and MAE criterions.

Results: When we look at the model of accuracy measures, it can be said that Winter's exponential smoothing is more accurate than Box-Jenkins method. The multiplicative model of Winter's exponential smoothing method is preferred because the MAPE and MAE values are smaller.

Conclusion: According to multiplicative Winter's exponential smoothing method, it is predicted that commercial egg production will increase 3.27% in 2018, 7.69% in 2019 based on the year 2017. According to results, the amount of commercial egg production tends to increase between 2018 and 2019. Forecasting commercial egg production can be used in the poultry sector and useful for better decision making and production planning for the future.