Year: 2025
Type: article
Abstract: Current stock market price prediction research mainly uses machine and deep learning for historical data, sentiment, and macro - indicators to boost accuracy. Prediction is crucial for investors, risk management, and market stability. LSTM has strengths like handling long - term dependencies in price sequences but has long training times and high resource use. This paper gets index volatility and return data via GARCH, then uses Bayesian optimization on LSTM to enhance prediction. It validates the model by comparing four metrics with others. Using the CSI 1000 Index, the Bayesian - optimized LSTM reduces RMSE by 0.169% compared to the basic LSTM.
Author: Chen Ting
Institution: Shanghai University
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W4410414180