Agricultural product price prediction based on compressed sensing and deep learning
In this paper,a combined model that combines compressed sensing(CS)and deep learning is adopted.The orthogonal matching pursuit(OMP)algorithm is selected from the compressed sensing algorithm,and the long short-term memory(LSTM)model is selected from the deep learning model.Based on the original agricultural product price data that may have noise in recent years,the original data are denoised by sparse representation,design of observation matrix,signal reconstruction and other steps,and then the reconstructed data after noise reduction and LSTM model are combined to predict the price trend of agricultural products.The combined model has the advantages of low requirements for data storage and low sensitivity to small noise.Compared with the traditional support vector regression(SVR)model,the accuracy of the prediction data is about 13％higher.Compared with other models,the combined model also has higher accuracy,and can achieve more accurate prediction results than the traditional models within a one-year time span.
NETL
NSTL
万方数据
