Stock market have always been a subject and a matter of immense interest due to it’s bumpy and volatile nature. Researchers have made attempts to ameliorate the existing models to better predict the share prices. This paper presents a comparative performance analysis of enhanced Long Short-Term Memory (LSTM) models for stock price prediction. The study explores two approaches: one involving hyperparameters tuning via Optuna and the other adding an error correction layer. Stock price data for Adani Enterprises from January 2012 to May 2024 is used, with the closing price as the output variable. The LSTM model’s effectiveness is evaluated using statistical metrics like Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). This work contributes to the ongoing efforts to enhance LSTM-based models for time series forecasting particularly in the volatile domain of stock market.