Stock fluctuation prediction method and device, computer equipment and storage medium

Abstract

The invention discloses a stock fluctuation prediction method and device, computer equipment and a storage medium. The method comprises the steps of obtaining historical data of a target stock and atleast one group of related data influencing fluctuation of the target stock; calling an unsupervised learning model to learn the at least one group of related data to obtain a volatility implicit state set corresponding to the at least one group of related data; obtaining volatility implicit state transition probability information according to the volatility implicit state set; determining a target stock fluctuation rate explicit state corresponding to the fluctuation rate implicit state set based on the historical data; and determining the fluctuation trend of the target stock according to the fluctuation rate hidden state transition probability information and the target stock fluctuation rate display state corresponding to the fluctuation rate hidden state set, thereby improving the accuracy of predicting the stock fluctuation trend.

Description

technical field [0001] The present application relates to the field of computer technology, and in particular to a stock fluctuation prediction method, device, computer equipment and storage medium. Background technique [0002] At present, stock volatility prediction methods in the market mainly use time series analysis based on the historical return standard deviation and logarithmic return rate of stock prices, such as GARCH (Autoregressive Conditional Heteroskedasticity Model, autoregressive conditional heteroscedasticity model), Models such as ARMA (Autoregressive Moving Average Model) predict the specific value of the volatility of the stock return rate. [0003] There is a major flaw in the prediction of traditional GARCH, ARMA and other models, that is, it fails to consider that the stock return rate is non-stationary, and only analyzes the historical return standard deviation and logarithmic return rate of the stock price to predict. The volatility trend of the sto...

AI Technical Summary

Problems solved by technology

[0003] There is a major defect in the forecasting of traditional GARCH, ARMA and other models, that is, it fails to take into account that the stock return rate is non-stationary, and only analyzes the stock price historical return standard deviation, logarithmic return rate and other characteristics to predict The volatility trend of the stock, the accuracy of the prediction result is not high

Images