NORMA eResearch @NCI Library

Prediction Stacking for long-term forecasting of the Bombay Stock Exchange Index by leveraging Machine and Deep Learning Models

Rao, Sumanth Bijadi Sridhar (2020) Prediction Stacking for long-term forecasting of the Bombay Stock Exchange Index by leveraging Machine and Deep Learning Models. Masters thesis, Dublin, National College of Ireland.

[img]
Preview
PDF (Master of Science)
Download (3MB) | Preview

Abstract

Stock markets are considered to be very volatile and for the most part unpredictable. It is certainly challenging to predict stock indices on a short-term basis let alone forecast it over a long period of time. In this research project, the main objective is to find reasonable answers to the research questions considered and to do so, the prediction model stack approach is proposed in order to combine the predictive capabilities of the individual machine and deep learning models to further optimize the predictions made, while, significantly reducing the margin of error. Results obtained showed that the BSE stock index can be predicted to a great extent over a long period of time and that the LSTM RNN model predicts the future stock index prices quite accurately and it achieved an R2 score of 0.9492. The RF-DLSTM prediction model stack, implemented using the proposed prediction model stacking approach, outperformed the LSTM RNN model with an R2 score of 0.966, RMSE score of 0.0503 and MAE score of 0.0408.
Keywords: Financial Forecasting, Regression, Stock Index, Custom Loss Function, Deep Neural Network, Long-Short Term Memory network, Decision Tree, Gradient Boosting, Random Forest, Prediction Stacking, Forecasting, Machine Learning, Deep Learning.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science

Q Science > QA Mathematics > Computer software
T Technology > T Technology (General) > Information Technology > Computer software

H Social Sciences > HG Finance > Investment > Stock Exchange
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Dan English
Date Deposited: 23 Jun 2020 10:44
Last Modified: 23 Jun 2020 10:44
URI: http://trap.ncirl.ie/id/eprint/4313

Actions (login required)

View Item View Item