Predicting Stock Price by Bidirectional-Lstm with Dual Attention Network

Predicting stock prices is one of the challenges in the field of financial markets.Due to the high volatility of stock prices, traditional data prediction methods cannot identify the most relevant and critical market data for predicting stock prices.This study discusses the use of Bidirectional-LSTM and Dual Attention Network in the process of predicting stock prices using stock data of state-owned banks.from historical Yahoo finance.The purpose of this study is to build a model that applies the two methods of Bidirectional-LSTM and Dual attention Network.The features used in the prediction process are open, high, low, volume, adj close.The method proposed by this research is compared with the LSTM and BI-LSTM models.The results show that the performance of the proposed method is the best, MAE, MSE, RMSE are the smallest with the results MAE = 0.079, MSE = 0.012, RMSE = 0.110.