Foreign exchange rate prediction method based on combination of neural network model and classification model

Abstract

The invention discloses a foreign exchange rate prediction method based on combination of a neural network model and a classification model, and relates to the technical field of data processing methods of foreign exchange rate prediction systems. The technical defects that a foreign exchange rate prediction scheme is large in uncertainty and high in trial and error cost, and the model dissimilarity is not fully utilized are overcome. For the regression model, performing data preprocessing on the foreign exchange historical data in the N periods, and inputting the data into a neural network model; adding a delay layer into the neural network model, and performing iterative prediction to obtain an optimal prediction result; for the classification model, collecting foreign exchange historical data in the same period as the regression model, and dividing a confidence interval by taking price data predicted by the regression model; predicting a price data rising and falling probability value according to the classification model, and adjusting the size of a confidence interval; therefore, a suitable foreign exchange rate prediction result for price reference of entrance and exit is obtained. The influence of uncertainty is reduced, and the reliability of a prediction result is improved.

Description

technical field [0001] The invention relates to the technical field of a data processing method of a foreign exchange rate prediction system. Background technique [0002] With the development of productivity and the globalization of finance, economic exchanges between countries in the world have become more and more frequent, further promoting the integration of the global economy. The fluctuation of foreign exchange rate not only has a relatively far-reaching impact on international financial relations, but also affects the economic development of many fields in the country, playing an increasingly important role. As far as enterprises are concerned, many enterprises are involved in transnational business. In addition, many financial, Internet and other enterprises are engaged in some securities and futures investment work. The foreign exchange market contains high leverage, which contains internal With the characteristics of high risk and high profit, how to accurately p...

AI Technical Summary

Problems solved by technology

[0013] To sum up, the purpose of the present invention is to solve the existing foreign exchange rate forecasting scheme for foreign exchange historical data, which has large uncertainties, high trial and error costs, and lack of technical deficiencies in making full use of the dissimilarity of the model, and proposes a Forecasting method of foreign exchange rate based on the combination of neural network model and classification model

Images