Abstract:
Cassava is an important cash crop in Thailand. Thailand ranks the worlds second biggest producer of cassava and also the major exporter of cassava and related products. The objective of this research by study the factors that affect cassava yield, and comparing the efficiency between multiple regression analysis and artificial neural network. This study uses secondary data of the year 2019 from 313 districts in the northeastern part of Thailand having cassava cultivation, collected by various government agencies. By comparing the efficiency of prediction methods with mean squares error, results show that multiple regression analysis gives mean squares error of 13,334.57, artificial neural network method using the program R-Studio gives mean squares error of 12,286.38. In conclusion, the artificial neural network method using the program R-Studio is more effective in predicting cassava yield in the northeastern region of Thailand than by multiple regression analysis.