NORMA eResearch @NCI Library

Sector Based Stock Market Prediction In USA

Uddin, Muhammad Nizam (2021) Sector Based Stock Market Prediction In USA. Masters thesis, Dublin, National College of Ireland.

[img]
Preview
PDF (Master of Science)
Download (4MB) | Preview
[img]
Preview
PDF (Configuration manual)
Download (7MB) | Preview

Abstract

At a time when stock market investments have seen a surge and market unpredictability has hit new heights with global recessions due to events such as the covid 19 pandemic of 2020, this research offers a fresh outlook to lure in new investors and provide a safety net for seasoned investors by implementing machine learning models to predict stock market behaviour. This has been accomplished by building an Sector based index to segment out the companies and aggregating the closing prices based on that to determine how the Sector would perform over time using RNN, LSTM and Time Series ARIMA. Among the model LSTM obtained the average highest-RMSE value 6.64 with model accuracy 93.65%. This has been attained by establishing indices not just for companies but also paving the path for sector based segregation to enhance the research and make these indices more accurate over time. The research provides a crisp comparison between the existing models and highlights the pros and cons of implementing a new model as proposed for the research.

Item Type: Thesis (Masters)
Subjects: E History America > E151 United States (General)
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: Clara Chan
Date Deposited: 15 Dec 2021 11:01
Last Modified: 15 Dec 2021 11:01
URI: http://norma.ncirl.ie/id/eprint/5231

Actions (login required)

View Item View Item