Document Type



Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence


Computer Sciences

Publication Details

A dissertation submitted in partial fulfilment of the requirements of Technological University Dublin for the degree of M.Sc. in Computer Science (Data Science), 2021.


The volatility of stock markets makes them notoriously difficult to predict and is the reason that many investors sell out at the wrong time. Contrary to the efficient market hypothesis (EMH) and the random walk theory, contribution to the study of machine learning models for stock price forecasting has shown evidence of stock markets predictability with varying degrees of success. Contemporary approaches have sought to use a hybrid of convolutional neural network (CNN) for its feature extraction capabilities and long short-term memory (LSTM) neural network for its time series prediction. This comparative study aims to determine the predictability of stock price movements by using a hybrid convolutional neural network (CNN) and long short-term memory (LSTM) neural network, a standalone LSTM neural network, a random forest model, and support vectors machines (SVM) model. Specifically, the study seeks to explore the predictive ability using stock price data, technical indicators, and foreignexchange (FX) rates transformed into deterministic trend signals as features for a hybrid CNN-LSTM neural network. This paper additionally considered including news article sentiment scores relating to stocks as part of the training dataset, but significant correlation was not found. In this study, the predictive ability is the accuracy of predicting the direction a stock price moves not the actual price.

The experiment results suggest that a hybrid CNN-LSTM model can achieve around 60% accuracy trained with deterministic trend signals for stock trend prediction. This accuracy has higher than the accuracy of LSTM, random forest, and SVM. On this basis, one can conclude that the hybrid neural network model is superior to standalone LSTM, random forest, and SVM for stock price trend prediction.