Data Mining Techniques for Predicting Cassava Yields in Lower Northern Thailand
Anamai Na-udom, Jaratsri Rungrattanaubol · Journal of Telecommunication Electronic and Computer Engineering (JTEC) · 2017
This paper investigates the factors influencing the cassava yields and develops the predictive models to predict the cassava yields in lower northern Thailand. The main objective is to compare the prediction accuracy between data mining technique namely Artificial neural network model and the conventional model namely Stepwise regression model. The root mean square error and mean absolute error values are used to validate the prediction accuracy. The results show that the significant factors are plantation area, cassava variety, cultivation period, and quantity of fertilizer. Further Artificial neural network performs better than stepwise regression model in terms of prediction accuracy. The results obtained from this study will assist farmers to improve their practices in order to increase the cassava yields.