Agricultural economy prediction method based on improved BP neural network

This paper deals with the problem of agricultural economy prediction method, which is a crucial problem in modern rural economy management. The main innovations here is an improved Back Propagation (BP) neural network which is exploited to solve the agricultural economy prediction problem. The BP neural network utilizes the algorithm of multilayer neural network model. As the BP algorithm cannot effectively search the global optimum solution of the prediction application, in this paper we develop an improved BP neural network based on genetic algorithm to promote the accuracy of state prediction. The genetic algorithm refers to a policy to solve both constrained and unconstrained optimization problems using a natural selection process. Additionally, the proposed algorithm repeatedly modifies a population of individual solutions, and then suitable individuals are chosen by the genetic algorithm from the current population. To demonstrate the effectiveness of the proposed algorithm, we choose the agricultural economy data from China’s statistical yearbook to construct two datasets, which are corresponding to grain yield and index of gross agricultural output value respectively. Experimental results verify that compared with other methods, the proposed algorithm can achieve higher prediction accuracy and more stable prediction ability. © 2016, UK Simulation Society. All rights reserved.