Security Risk Analysis and Price Predictions with Machine and Deep Learning Models (LSTM)

Muhammad Mehdi Muhammad Mehdi,Inam Ul Haq Inam Ul Haq,Atif Ikram Atif Ikram
DOI: https://doi.org/10.56536/jicet.v4i1.103
2024-03-30
Abstract:Risk analysis and price predictions of securities, shares and stocks, have been a challenging problem for investors. Many factors, Economic, Political etc., can disturb stock returns and their prices. However, with the advent of Artificial Intelligence and Machine/Deep Learnings techniques, predictions of returns and prices have become very easy with accuracy and precision. In this research paper focus is given to Long Short Term Memory (LSTM) model, closing prices of Tech stocks/shares of companies in technology industry are predicted, RMSE and accuracy (MAPE) is calculated and then compared these values with the predicted and calculated values of Feedforward Neural Network (FNN) model and Recurrent Neural Network (RNN) model. After using weighted average method, LSTM is proven to be one of the best models to predict securities prices and returns.
What problem does this paper attempt to address?