Empirical mode decomposition and ANFIS network-based prediction technique for financial forecasting.

A financial market is non-linear and chaotic in nature. So, the accurate prediction of foreign exchange rate is very difficult and challengeable task. Hence, many proposed techniques and new approaches are used for forecasting various countries' exchange rates with different parameters. This paper proposes EMD-ANFIS for foreign currency exchange rate prediction. In this research, we would like to propose a model which could develop multivariate exchange rates information and put these features to better use. The performance of the proposed system has been tested with European EURO against US Dollar (EUR/USD), British POUND against US Dollar (GBP/USD), US Dollar against Swiss FRANK (USD/CHF), US Dollar against Japanese YEN (USD/JPY) and used to predict one day exchange rate in advance. Empirical mode decomposition (EMD) and QPSO (Quantum Particle Swarm Optimization) are techniques that used here which generates optimal weight for the proposed model. The proposed approach has been found with the best prediction rate against previous studies. [ABSTRACT FROM AUTHOR]