Research And Application Of Fuzzy Similarity And Grey Model In Prediction Of Stock Price Turning Point

Stock price prediction is a key application in the application of stock investment.Accurate prediction of the trend of the stock price is the core reference for the decision-making stock investment.Aiming at the issue of stock price prediction,this paper takes the grey theory as the main line,and then constructs a variety of adaptive grey models with the help of error feedback mechanism,residual correlation coefficient,genetic algorithm and other methods to realize the effective prediction of stock price.After analyzing the basic factors of stock,the mathematical framework of time series reconstruction of stock price inflection point is constructed by mathematical modeling based on the entanglement theory,which is used to realize time series reconstruction of stock price that is for the verification of subsequent models.By using the error feedback mechanism of control theory to minimize the average relative error,we put forward an error feedback grey model that can find the optimal parameter array automatically and avoid the problem that the matrix inverse is not available.This model is for the prediction of stable stock price.By considering the spatial structure information of time series,a residual correlation coefficient is proposed,based on which a similar grey model is therefore designed.This model makes full use of the similar historical data segment information to realize the current stock price prediction.In order to make up for the shortcomings of ergodic optimization of error feedback grey model and similar grey model,a genetic optimization grey model is proposed based on genetic algorithm,and the fuzzy expression of prediction results is realized based on neighborhood relationship and fuzzy operator.In the experimental verification part of all the proposed models,the most commonly used three stock time series are taken as the research object,and a variety of related grey models are compared to verify the effectiveness of the models proposed in this paper.