Stock Market Prediction Using Deep Attention Bi-directional Long Short-Term Memory

Trustworthy predictions of future stock can promote significant profits, and it has attracted several financial analysts and investors. Accuracy suffers when more features are added and time consumption increases. To address these issues, this study proposes an Effective Stock Market Prediction with a Deep Attention BiLSTM framework optimized utilizing the COOT Birds Algorithm (DABiLSTM-COOT). Initially, the stock data are collected from the NSE stock dataset (Nifty 50), and the technical and fundamental indicators are measured for effective closing price prediction. The optimal features are chosen by adopting the Improved Binary Butterfly optimization (IBBO) algorithm. The DABiLSTM-COOT method predicts closing prices more accurately. Also, the performances are analyzed in terms of precision, recall, and F1 score, MSE, RMSE and MAPE.