Prediction of wheat stripe rust based on support vector machine

Wheat stripe rust is the most important wheat disease in China. The prediction of this disease is significant to making control strategies and taking timely management measures to ensure high and stable yield of wheat. Compared with regression analysis method, support vector machine method was be used to predict wheat stripe rust. The results showed that support vector machine method can achieve higher fitting accuracy and prediction accuracy, and that application of this method for the prediction of wheat stripe rust is feasible and efficient. In practical applications, optimal combination of SVM types and kernel functions can be selected to carry out the disease prediction. Therefore, a new approach was provided for the prediction of wheat stripe rust.