LSTM-based cross-prediction price model for gold and bitcoin

Since the rise of Data Analysis, forecasting of price markets has never stopped and there are numerous forecasting methods, but most of them are only for a single price data.We have chosen bitcoin and gold as the subjects of our study, addressing the multi-objective related prediction problem, explores the volatility relationship between gold and bitcoin to improve its forecasting accuracy, and in doing so, we establishes multiple prediction models,and determines the relationship between prediction accuracy and prediction range. We first performed the gray correlation analysis and the wavelet coherence analysis on the market data of bitcoin and gold to exam the time-frequency structure of correlation and co-movements between the gold futures and bitcoin markets.We found that there is a relatively high co-movement between gold and bitcoin in the frequency band from 2018 to 2021, and a lag of about 4 weeks of bitcoin to gold stock price. Based on this finding, a many-to-many LSTM model was built with an accuracy of 0.79 by parameter search. In addition, to further corroborate the accuracy of the LSTM, using RMSE as a criterion, we also built a support vector machine, Gaussian regression, time series, and simple regression tree models.