Prediction of grain yield based on spiking neural networks model

Grain yield is important in national economy so there is necessary for grain yield prediction. A novel predicting model based on spiking neural networks (SNNs) is presented for this purpose. SNNs are computationally more effective than conventional artificial neural networks. The spiking neurons act as basic elements in which information deliver from one neuron to another in forms of multiple spikes via plenty of synapses. Besides, the corresponding learning mechanism called Spikeprop is also discussed. An example, prediction of China annual grain yields as our experiment, is used to explain the principle of SNNs based method. Experimental results are demonstrated showing the feasibility and accuracy of our approach.