This study used neural networks to evaluate the incidence of anthracnose of techno's chili in Chiang Mai, the data collected by the National Weather Service in Chiang Mai last four years, a total of 1096 sets of data share data. two-part The first part will bring back four years to train 600 series models of neural networks. The input variables include the maximum temperature. And maximum relative humidity The data for the remaining 496 data sets to test the prediction. And measure the performance of the prediction using the square root of the mean square error average (root mean square error, RMSE) and the acceptance (agreement of index, IA), the study found that the model neural network there. very effective Have predicted quite accurately In terms of a prediction, there are problems with the data. (Pre-process) should increase accuracy by increasing the number of data sets, learning neural network model. The program allows for effective monitoring or surveillance of the disease, Ann Tracks Aires more.
การแปล กรุณารอสักครู่..