Wavelet Denoising and Attention based RNN ARIMA Model to Predict Forex Price

This project I supervised in Semester 2, 2019. A paper has been accepted in IJCNN 2020 and soon will be available in IEEE Xplore. In every trend change of theforex market there is a great opportunity as well as risk for investors. Accurate forecasting of forex pricesis crucial element in any effective hedgingor speculation strategy. However, the complex nature of the forex market makesthe forecasting problemchallenging, which has prompted extensive research from various academic disciplines. In this paper, a novel approach that integrates the wavelet denoising, attentionbased recurrent neural network (ARNN), and autoregressive integrated moving average (ARIMA)is proposed. Theoretically, the wavelet transform could remove the noisefrom the original time series to stabilize the data structure. Meanwhile, ARNN model is flexible to capture the robust and nonlinear relationships in the sequence and ARIMA can well fit the linear correlation of the sequential information. By