Financial Fortunes: A Stock Price Prediction Using LSTM

This paper uses a variety of libraries and tools to present a comprehensive stock price prediction and sentiment analysis tool. Initialization is the first step, after which data preprocessing and library imports are completed. Tensor Flow and Keras are used in the construction of the LSTM model, while the Adam optimizer and mean squared error are used in training. The script reads the actual stock prices, scales it for prediction and then makes out predictions using testing data. Also, it uses Matplotlib to visualize real and predicted stock prices as well as computing R-squared accuracy score. The script also applies Long Short-Term Memory (LSTM) neural networks along with social media sentiment analysis from a designated subreddit in developing a stock price prediction model that appears quite promising with 94.1% accuracy.