Predicting Quantity of Coffee Consumption in China Using Grey Prediction with Fourier Series

Man-Ching Peng ； Yi-Chung Hu ； Hang Jiang

《Journal of Grey System》 20卷3期 (2017/09) Pp. 163-169

摘要

The worldwide annual coffee consumption per capita is over two hundred cups in year 2015, whereas it is only four cups in China. With the increasing acceptance of coffee, China will be a great potential market for coffee consumption. Accurate prediction of China's coffee consumption is crucial for investors to select appropriate timing to enter into market. The GM(1,1) model has drawn out attention to coffee consumption forecasting because it only needs limited samples to construct a time series without statistical assumption. To further improve prediction accuracy of GM(1,1) model, this paper employs Fourier series to modify the residuals. It is found that the original GM(1,1) with Fourier series performs well from collected data compared to grey prediction models considered. So this model was further applied to predict China's coffee consumption for next decade. Based on the simulation results, this paper suggests that it is appropriate to enter the market before year of year 2020.

關鍵字

Coffee consumption ； China ； GM(1,1) ； Prediction accuracy ； Fourier series