A NOVEL SYSTEM FOR PREDICTING THE DAMAGE OF RICE DISEASES IN AN GIANG PROVINCE, VIETNAM

Abstract

Many researchers have recently considered information technology applications to support smart agriculture. Numerous solutions and information technology systems have been implemented to benefit farmers, such as an e-commerce website for exchanging agricultural products, an agricultural information system, a livestock information management system, and an agricultural environment monitoring system. This paper presents a novel system that combines artificial back-propagation neural networks with genetic algorithms for predicting the damage of rice diseases in An Giang province. The system predicts rice disease damage in the coming weeks in 11 districts in An Giang based on input parameters such as climate conditions and previous rice disease damage.  The approach was evaluated on a dataset collected by the An Giang Plant Protection Department over 22 years. The rice diseases forecasting system produced promising results and received positive feedback from forecast department specialists, with an RMSE of 19.31 on the test dataset (ha).