Research Article
The Foreign Exchange Asset Pricing Model Deeply Integrating ARIMA with Decision tree and LSTM
@INPROCEEDINGS{10.4108/eai.2-12-2022.2328736, author={Jiankun Sun and Xin He and Weijie Zhang and Tianjiao Zhao}, title={The Foreign Exchange Asset Pricing Model Deeply Integrating ARIMA with Decision tree and LSTM}, proceedings={Proceedings of the 3rd International Conference on Big Data Economy and Information Management, BDEIM 2022, December 2-3, 2022, Zhengzhou, China}, publisher={EAI}, proceedings_a={BDEIM}, year={2023}, month={6}, keywords={inflation rate decision tree model genetic algorithm arima model lstm model}, doi={10.4108/eai.2-12-2022.2328736} }- Jiankun Sun
Xin He
Weijie Zhang
Tianjiao Zhao
Year: 2023
The Foreign Exchange Asset Pricing Model Deeply Integrating ARIMA with Decision tree and LSTM
BDEIM
EAI
DOI: 10.4108/eai.2-12-2022.2328736
Abstract
In recent years, with the improvement of economic development level, China's consumer price index has increased year by year. The central bank faces huge inflationary pressure. Many domestic scholars have done a lot of research on the simulation and prediction of inflation behavior. On the basis of previous research results, we selected econometrics, traditional machine learning, deep learning three typical sequence prediction model, to explore their scope of application, using ARIMA, LSTM, decision tree model of China inflation short-term forecast, and according to "sample size preference" and "fluctuation intensity preference" as the characteristics of performance comparison, the results show that the decision tree model prediction effect is better than ARIMA, LSTM inflation prediction model. In view of this, we propose that the decision tree model be used more broadly in the field of inflation rate prediction, thus providing a more valuable reference for formulating macro policies.
