Optimized Artificial Intelligence and Econometric Model Empowered Virtual-Fiat Settled Price Prediction in Green Cryptocurrency Networks
Xiaotong Jiang, Jun Wu, Qianqian Pan · 2024
Recently, the promotion of green cryptocurrencies has attracted attention due to the huge resource consumption brought by cryptocurrency transactions. The main driver of green cryptocurrencies is to reduce resource consumption and reduce the number of transactions. Accurate prediction of cryptocurrency prices is difficult because they are influenced by diverse kinds of factors besides supplydemand relationship. First, the price of cryptocurrency changes rapidly and fluctuates violently, so the traditional econometric methods cannot respond well to the drastic price changes in a short period of time. Second, artificial intelligence (AI) models are relatively separated from econometric cryptocurrency price prediction models, which leads to deviations when forecasting in the financial field. Third, existing AI models have not been well optimized specially for cryptocurrency prediction. To address the above challenges, in this article, we propose the optimized AI and econometric model to empower virtual-fiat settled price prediction in green cryptocurrency networks. In our proposed econometric model renew autoregressive integrated moving average (REARIMA), the problem of poor econometric model response to drastic changes in a short time is solved by the joint design with AI. Moreover, the AI model innovation optimization for the prediction of cryptocurrency is carried out in dense long-short term memory (DENSE_LSTM). Finally, DENSE_LSTM are used to optimize the econometric model. The feasibility of the proposed model is verified by experiments.