Among all diseases affecting rice production, rice blast disease has the greatest\nimpact. Thus, monitoring and precise prediction of the occurrence of this disease are important;\nearly prediction of the disease would be especially helpful for prevention. Here, we propose an\nartificial-intelligence-based model for rice blast disease prediction. Historical data on rice blast\noccurrence in representative areas of rice production in South Korea and historical climatic data are\nused to develop a region-specific model for three different regions: Cheolwon, Icheon and Milyang.\nA rice blast incidence is then predicted a year in advance using long-term memory networks (LSTMs).\nThe predictive performance of the proposed LSTM model is evaluated by varying the input variables\n(i.e., rice blast disease scores, air temperature, relative humidity and sunshine hours). The most\nwidely cultivated rice varieties are also selected and the prediction results for those varieties are\nanalyzed. Application of the LSTM model to the accumulated rice-blast disease score data confirms\nsuccessful prediction of rice blast incidence. In all regions, the predictions are most accurate when all\nfour input variables are combined. Rice blast fungus prediction using the proposed LSTM model is\nvariety-based; therefore, this model will be more helpful for rice breeders and rice blast researchers\nthan conventional rice blast prediction models.
Loading....