Enhancing Price Prediction Accuracy through LSTM and Comparative Deep Learning Models

Risk analysis and price predictions of securities, shares, and stocks have been a challenging problem for investors. Many factors, including economic and political ones, can affect stock returns and their prices. However, with the advent of Artificial Intelligence and Machine/Deep Learning techniques, predictions of returns and prices have become easier with increased accuracy and precision. This research paper focuses on the Long Short-Term Memory (LSTM) model, which predicts the closing prices of Tech stocks/shares from companies in the technology industry. The RMSE and accuracy (MAPE) are calculated. These values are then compared with the expected and computed values of the Feed forward Neural Network (FNN) model and the Recurrent Neural Network (RNN) model. After using the weighted average method, LSTM has been proven to be one of the best models for predicting securities prices and returns.