Machine Learning Models for Financial Applications

Abstract

The prediction with high accuracy and speed is one of the most important tasks in machine learning to get the trend and results for the given dataset. Machine learning is used in different fields and one of these fields is the finance field so one of the applications of machine learning in finance is predicting stock market value and classifying its direction. The stock market is a platform where individuals are purchasing and vending shares of publicly traded companies in it with main goal which is making money. The performance of the stock market will have an impact upon economic growth. Thus, the goal is to maximize the profit and minimize the losses so machine learning techniques are used to predict the stock market values and classify its direction with high accuracy by gathering the historical dataset over the last few years. This thesis contribution is to outperform the accuracy result of stock market prediction value and its direction than other new published studies and papers that will be listed by comparing the accuracy and R-squared results. The proposed solutions are divided into two solutions by applying a neural network. The first proposed solution is classifying stock market direction into positive and negative indicators by applying artificial neural network architecture. The second proposed solution is predicting stock market value by applying long short-term memory architecture.