A comparison of artificial neural network and regression model for predicting the rice production in Lower Northern Thailand

Abstract

Lower Northern Thailand is one of the main regions which can produce the highest rice yield. If the emphasis is on producing the rice yield in order to meet the standard yield, then the key factors, such as characteristics of rice farm, rice seed types, cultivation period, quantity of fertilizer usage, number of seeds, must be clearly studied and understood. This paper studies factors influencing the rice products and develops a model to predict rice yield per rai that can support farmers to plan their rice farming in Lower Northern Thailand. The aim of this paper is to compare the prediction accuracy between two popular predictive techniques for modelling rice yield namely, artificial neural network (ANN) and Regression. Root mean square of error (RMSE) and mean absolute error (MAE) values are used to compare prediction accuracy of the predictive models. The result shows that ANN is superior over regression model in terms of prediction accuracy and it is flexible to develop.