An efficient deep learning algorithm for predicting market trends using cyclic long short-term memory network

Abstract

The stock market's role in the economy has attracted researchers. Many researchers analysed stock market trend and price prediction. Previous study used neural network and statistical models to predict experimental results. Deep learning has a tremendous learning capacity and is appropriate for complicated financial time series. The cyclic long short-term memory (LSTM) network is well-suited for theoretical financial time series prediction. This study proposes an efficient deep learning approach for stock market trends prediction. This deep learning framework includes data processing, deep learning models, and prediction optimization. An optimizer namely whale optimization algorithm (WOA) enhance the RNN-LSTM network-based deep learning network prediction. Comparative models demonstrated that the proposed framework is efficient. Testing and data analysis showed that the proposed framework is effective at predicting stock market trends.