Evaluating the Inclusion of Technical Indicators for Deep Learning-Empowered Stock Price Prediction in the Chinese Market
摘要
There have been extensive research attempts to predict stock prices or movements utilizing deep learning models such as Long Short-Term Memory (LSTM)-based models, and many have demonstrated potential results. In recent literature, technical indicators such as RSI, WR, BIAS, and KD added as part of the input data have been explored and some reported improved performance with the inclusion of these indicators. However, most of such studies were in the context of the US market and they have reported the positive results of including technical indicators but have not given much explanation. In this paper, we compare various LSTM-based models and evaluate the impact of incorporating technical indicators in the context of the Chinese stock market. The experimental results answer key questions regarding the consistency of benefits from technical indicators in a less mature and developing stock market. The findings indicate that the inclusion of technical indicators could offer some benefits by a small margin for the Chinese market. Finally, several potential future research avenues are highlighted.