Combining technical analysis and deep learning models for stock market trading

Abstract

The issuance of stocks constitutes a means by which ownership in a company is represented, and its distribution may vary depending on whether the company is limited or public. The stock market offers the potential for high returns, thereby serving as an attractive avenue for investment. Against this backdrop, the objective of this study is to develop a predictive model for stock prices that can facilitate profitable trading outcomes. To achieve this aim, the study focuses on intraday and hourly trading and utilizes a hybrid model that integrates Bidirectional Long Short-Term Memory (BiLSTM) and Convolutional Neural Network (CNN) architectures, along with technical indicators. BiLSTM is a neural network architecture that possesses the capability to process sequential data in both forward and backward directions, thereby augmenting the model's ability to capture dependencies within the data. The efficacy of the resulting model is subsequently evaluated through a comparison with technical analysis. Empirical validation of the model is carried out using technology stocks that are listed on the NASDAQ index. The experimental findings demonstrate that the hybrid architecture of CNN and BiLSTM can outperform technical analysis in terms of achieving profitable trading outcomes in the stock market.