Asian Journal Of Social Pharmacy

Abstract: Objective To predict the price of Chinese herbal medicines and to provide references for the government policy making and enterprises operation based on a case study. Methods ARIMA model and GM (1, 1) model were used respectively to predict the price of Panax Notoginseng, then variance reciprocal method was used to test the results of the combined prediction. Results and Conclusion GM (1, 1) model is better than ARIMA model to predict the price, but the combination model is the best. Combination forecasting model is more accurate to predict the price of Chinese herbal medicines.

Key words: price of Chinese herbal medicine, combination model, ARIMA model, gray GM (1, 1) mode