Automation of Financial Capital Supply Chain Optimization Industry Based on Time Series LSTM Algorithm

With the development of global economic integration and the increase of economic uncertainty in the future, it is very necessary to predict the future capital of financial supply chain industry in order to ensure that capital optimization is within a controllable range in the current macroeconomic environment and current situation. This paper proposed an automated capital prediction strategy of capital supply chain using time series analysis artificial intelligence methods. First, to analyze the fluctuation and tail risk of financial characteristic, this paper explores the financial characteristic to measure the dynamic VaR from the perspectives of volatility, tail and peak with the Bayesian peaks over threshold (POT) model. Then, in order to make the modeling more refined, the forecast targets are split before the modeling with seasonal Autoregressive Integrated Moving Average (ARIMA) model and Prophet model. Finally, the time series modeling of wavelet Long Short-Term Memory (LSTM) model is carried out by two-part analysis method to get linear separated wavelet and non-linear embedded wavelet parts to predict the strong volatility financial capital. Taking the user capital flow of Yu’e Bao platform, the results prove the feasibility and prediction accuracy of the innovative proposed model.