Research on Predicting the GBP/USD Exchange Rate Based on BP Neural Network
Z. Wu, Yihan Xie · Transactions on Computer Science and Intelligent Systems Research · 2024
As global base currencies, the fluctuations in the prices of the pound and dollar have an impact on the global economic situation. The exchange rate of currency not only affects the decline of the economy but also affects the stability and development of political and trade relations in various countries. Due to the numerous and non-linear factors that affect the currency exchange rates between the two countries, the accuracy of existing exchange rate prediction research is still relatively low. This experiment conducted a study on the prediction of the pound to dollar exchange rate, to help deepen the understanding of the operating rules of the international economy and financial markets and take corresponding measures. This experiment selected data on the GBP/USD exchange rate from May 12, 1993, to March 1, 2024, and used MATLAB to establish a Back Propagation (BP) neural network model to study the GBP/USD exchange rate. This article optimizes the prediction performance of exchange rates in this experiment by changing the number of hidden layer nodes as much as possible. The results show that the BP neural network in this experiment has a relative prediction error of less than 0.35 for the pound-to-dollar exchange rate, and has a good fitting effect.