Author(s)

Jiale Dong

Corresponding Author

Jiale Dong

ABSTRACT

This study proposes a stock price prediction model based on the hybrid LSTM-GAN (Long Short Term Memory Generative Adversarial Network) algorithm, combined with a multi factor risk quantification evaluation method, aiming to improve the accuracy and effectiveness of stock price prediction. By combining time series prediction with deep learning techniques, this study not only captures the nonlinear characteristics of stock prices, but also quantifies risk through multi factor models, providing strong support for investment decisions. In the experiment, the mean squared error (MSE) of the LSTM-GAN model was 0.2315, and the mean absolute error (MAE) was 0.3847. In the quantitative risk assessment experiment, the model predicted a portfolio Sharpe ratio of 0.2174 and a maximum retracement of 0.2351.In the above data conclusions, the superior performance of the hybrid LSTM-GAN model in stock price prediction and risk management is demonstrated.

KEYWORDS

Stock Price Prediction; LSTM-GAN Model; Multifactorial Risk; Quantitative Evaluation