Decomposition-based auto-regressive model for agricultural price prediction

Agricultural commodities prices are very unpredictable and complex; thus, forecasting these prices is one of the research hotspots. In this paper, a hybrid EMD-ARIMAmodel is proposed by combining empirical mode decomposition (EMD) with autoregressive integrated moving average (ARIMA) to improve the accuracy of agricultural price forecasting. Here empirical mode decomposition (EMD) adaptively represents non-stationary signals as sums of zero-mean amplitude and different frequency modulation components called intrinsic mode functions (IMFs) and residual. Then auto regressive integrated moving average (ARIMA) is constructed to individually forecast these IMFs and residual components. Finally, the prediction results of all IMFs, including residual, are aggregated to formulate an ensemble output for the original price series. Empirical results demonstrated that the proposed EMD-ARIMA model outperforms the ARIMA model regarding different forecasting evaluation criteria for international maize prices.