|
|
Predictability of yields of the main crops in Macedonia (FYROM)
Trajchevski, Hristijan
The objective of this thesis is to investigate yield predictability of the selected crops at national, regional and municipality level in Republic of North Macedonia. Emphasis is placed on wheat, maize, tobacco and grape, due to their significance for the economy of the country, both as exporting goods and for internal use. Specifically, the aim of the thesis is to investigate if there are possibilities for predicting the yield of the selected crops early in the season. The methods used in this thesis are yield estimation model based on linear regression using one variable and yield predicting model based on Artificial Neural Network (ANN) using meteorological data and remotely sensed data, individually. From the obtained results it is possible to conclude that the model based on linear regression indicates more accurate estimations for the yield of wheat and maize, than for the grape or tobacco. However, this model is not suitable for yield forecasting, whereas ANN is more suitable, because it manages to work with non-linear and very complex relationships. From the results of the yield predicting model it is possible to conclude that the model based on ANN using meteorological data showed more accurate results compared with the model based on ANN using remotely sensed data. Both of these model types generated highly accurate predictions, proving that ANN is powerful tool for yield prediction and forecast, even several months before the harvest.
|
guest ::
