Artificial Intelligence Based Hybrid Models for Prediction of Stock Prices

Abstract

Stock price forecasting has recently become an important practical component of the economic arena. An intriguing task, stock price forecasting is regarded to be related to the volatility and noise of stock market activity. To address these issues and accurately predict stock prices, this paper proposes a hybrid framework based on a learning model such as stacked Long Short Term Memory (LSTM) and Convolutional network. Experiments with several possible outcomes are run to assess the proposed framework using the stock price data set. The model was trained on ADANI stock price from the last roughly fourteen years on stacked LSTM with a Convolutional network and evaluated on an assessment criteria Root Mean Square Error (RMSE). The stacked LSTM model has proven to be a competitive model against the other models in stock price prediction in various scenarios.