The use of artificial neural network models to predict interest rates

Abstract

Using antipersonnel models of artificial neural network, (ANN) to gain access to interest rate as a new systematic science, proved successful technology in the financial area 0. This research aims to build models of neural network artificial which can increase the performance of prediction 'on corporate bond yields on corporate (CBY) as the rate of interest 0 analysis of market of synergies (SMA) was submitted to build models (Freedman) in this attribute 'while models that include only data of series time for crop Association of corporate has been developed, other models was generated through the link and the reorganization of key variables 0 per model was built to predict the 1.6 and 12 months as a period of time to predict. We got 9 models of artificial neural network to predict interest rate 0 artificial neural networks with multiple layers which used backpropagation (backpropagation) showed a good performance in the prediction for the period from month to six months.