The grey generalized Verhulst model and its application for forecasting Chinese pig price index

Weijie Zhou,Lingling Pei
DOI: https://doi.org/10.1007/s00500-019-04248-0
IF: 3.732
2019-08-01
Soft Computing
Abstract:The Verhulst model is used in many natural and social systems. When simulating and predicting system sequences with an inverted U shape or a signal peak feature by the grey Verhulst model (abbreviated as GVM), such as the pig price index, a drift phenomenon happens sometimes. We introduce a grey generalized Verhulst model named as GGVM to address this problem. Compared with the GVM, the GGVM contains a constant as the grey action quantity. Besides, the multivariable grey generalized Verhulst model is also given. Three cases containing Glipizide tablets blood drug concentration series, the number of traffic deaths, and consumer price index (abbreviated as CPI) sequences are utilized to demonstrate that GGVM can eliminate the drift phenomenon effectively. Given that the pork is the most consumed meat for Chinese residents and the fluctuation of its price is closely related to the interests of residents and pig-breeding enterprises, it is important to predict the pork price. So the prediction of pork price index whose time series possesses an inverted U shape is carried out by GGVM, GVM, and intelligent algorithm models, including LSSVR, ε$$\varepsilon $$-SVR, and RBF in the empirical part. The results show that GGVM produces higher accurate simulation and prediction than the models as given above.
computer science, artificial intelligence, interdisciplinary applications
What problem does this paper attempt to address?