| Abstract [eng] |
The dairy industry is of primary importance in Lithuania. There is about 32,5 thousand of dairy farms. Forecasting indicators of Lithuania dairy farms is a very important object. Many mathematical models and software were developed to predict the dairy farms economic indicators in the abroad. In Lithuania such models and software was not developed yet. The forecasting models for Lithuanian dairy farms milk production and milk procurement price is presented in this article. Models were developed using ANOVA, non-linear regression, time series methods and SAS system. Lithuania cows were setting into 23 homogenous groups and were developed nonlinear regression models of each milk production forecast. For these models correction using new data was choosed Levenberg-Marquardt algorithm. To forecast Lithuanian milk procurement price was developed seasonal trigonometric model. Milk production forecasting models were created into a website by using MS Visual Studio 2005 program. At first time was created a website for Lithuania farmers of forecasting milk production. |
