Local Linear Wavelt Neural Network Model for Prediction of Currency Exchnge Market

In this paper, Local Linear Wavelet Neural Network (LLWNN) – a combination of Artificial Neural Network and Local Linear Wavelet Technique- is proposed to predict the most chaotic financial time series data i.e. currency exchange rate. The gradient descent learning algorithm is used to train the proposed model. Indian Rupee against US Dollar is taken as the experimental data. A comparative result analysis between LLWNN model and Multilayer Perceptron (MLP) is also made. This model is used to predict currency exchange market for one day, one week and one month in advance. Tlie Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) are used to find out the performance of the LLWNN model. In case of LLWNN, both RMSE and MAPE are found to be less when compared with MLP model.