Research On Foreign Exchange Forecasting Based On Neural Network

Foreign exchange risk problem has attracted people's much attention and extensive research, and how to forecast the directions and the degree of the change of foreign exchange rate is the basis of foreign exchange risk management. At the same time, as an important index in international finance market and barbell that adjusts country economy internal and external balance, foreign exchange rate is deserved to research. This article discusses systematically current foreign exchange rate deciding theory,makes especial research on foreign exchange rate forecasting combined fashionable neural network technology now. First, this article discusses some kinds of foreign exchange rate deciding theory, and expatiates basic thought of different theories and makes some simple opinion. Second, makes out research progress of different foreign exchange rate forecasting technology, then discusses the principle and model of neural network in detail, and makes especial research on the model and algorithm of BP neural network, also puts forward better algorithm. At last, we apply neural network in foreign exchange rate forecasting system of industry and commercial bank Jiangsu province, design and implement neural network foreign exchange rate forecasting module and make forecast with actual data, also analyze on results. Experimental results demonstrate that the foreign exchange rate forecasting system can forecast effectively and we can achieve good test results with neural network.