Asian Research Journal of Current Science

The financial markets, being a very alluring innovation, have had a substantial impact on the economy of the country. Due to its high volatility and non-linear time data series, the stock market continues to draw traders and investors interested in engaging in simultaneous buying and selling. In this research, Long Short-Term Memory (LSTM) were used to create a model for predicting stock series. The research also investigates the viability of predicting stock market prices with the aid of historical data, trends, and machine learning algorithms. For this research, the data for Amazon stock prices were from 2011 to 2019 and Google stock prices from 2013 to 2017 which where gotten from the Kaggle Machine Learning repository. Several measures, including the R2 score, Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Mean Squared Error (MSE), were used in a comprehensive evaluation of the models used. The results of the experiments show that the Long Short-Term Memory (LSTM) prediction models perform better than the baseline models when it comes to making accurate forecasts. The Long Short-Term Memory (LSTM) model performed quite well, with an R2-Score of 0.9961 on the Amazon dataset and 0.9421 on the Google dataset. Taking everything into account, the results show that the Long Short-Term Memory (LSTM) model performed better than the state-of-the-art machine learning algorithms at predicting the values of Amazon and Google stocks.