International Journal of Scientific Research and Engineering Development

Stock price prediction remains an arduous undertaking due to nonlinear market patterns and volatility. This study develops a combined deep learning framework that integrates a Long ShortTerm Memory (LSTM) network with optimization principles derived from the Sparrow Search Algorithm (SSA). The concept in this paper stems from modern technology created to enhance prediction techniques based upon social media sentiment analysis and historical pricing data and technical indicators (moving averages, exponential moving averages, MACD and relative strength indexes). The combination of historical price data with technical indicators permits traders and investors to build predictive models for their respective trading methods. Historical price data and technical indicators represent the trends derived from time-series and technical analysis and allow analysts to make forecasts regarding future price levels. Testing results indicate that, in terms of Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE), models using Long Short Term Memory (LSTM) with recurrent networks that employ an optimization/addition technique outperform standard LSTM models. We have shown through our findings that optimization/addition-based recurrent neural networks (i.e., LSTM) offer a legitimate approach to short-term stock price forecasting given the rapidly evolving dynamics within the financial markets.

Dr. Manjunatha B N, Mrs. Nandini K V, M. Madhulika, P. Sravani, P. Likhitha, T. Harika,"Stock Price Prediction Using Deep Learning" International Journal of Scientific Research and Engineering Development, V9(3): Page(411-425) May-June 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.