Combination Forecast Model Based on Trend Decomposition and GIOWA Operators
Different data have their own different characteristics,for time series data,the data often have a time-related trend,by decomposing this trend and predicting the part outside the trend,we can get better prediction results.In this paper,the RMB/US dollar exchange rate data from January 4,2013 to August 11,2023 are selected,and the gray prediction model,ARIMA model,LSTM model and the combination of these models are used to predict the data,and the empirical results prove that the combined prediction model based on trend decomposition and GIOWA operator has better prediction accuracy than single forecasting.
combination forecastIOWA operatorsgray prediction modelARIMA modelLSTM model
NETL
NSTL
万方数据
