A hybrid LSTM driven predictive framework for Google stock price forecasting with market trend and volatility analysis

Accurate prediction of stock prices is important for the volatile and complex dynamics of the current market. Conventional statistical models are frequently unable to account for nonlinear relationships and time-series dependencies in financial time series, resulting in less accurate predictions. To solve this challenge, we propose a Hybrid LSTM-Driven Predictive Framework by incorporating market trend and volatility features along with historical price data to improve forecast accuracy. The base line includes technical indicators (SMA, EMA, RSI) and volatility measures (rolling standard deviation, GARCH) as well as sequential learning based on Long Short-Term Memory (LSTM) neural networks. Experiments conducted on Google stock price data illustrate that the hybrid model is more efficient than traditional machine learning models and the baseline LSTM when tested based on various evaluation metrics (MSE, RMSE, MAE, and R²). The results demonstrate that the LSTM + Trend + Volatility model obtains the minimum error values and the maximum R² (0.9911) compared to the baseline and state-of-the-art models. This study indicates that integrating trend and volatility data into sequential deep learning models enhances forecasting stability and accuracy for financial applications.