Capstone
During my Final Year I completed my Honours thesis which was titled Machine Learning for Stock Investing in the Technology Sector. I spent the first semester of 2023 conducting research related to my topic and formulating a research question and in the second semester actually working on the project and writing the report. The abstract for the report is below:
The main objective of this paper is to investigate the application of machine learning (ML) techniques for stock price prediction within the technology sector, in response to the burgeoning era of big data and technological progress. With a focus on Long Short-Term Memory (LSTM) networks within Artificial Neural Networks (ANNs), the study examines the capacity of machine learning to process and analyse datasets from 2016 to 2023, identify underlying patterns, and generate accurate stock price forecasts. A comparative analysis of predictions versus actual stock prices of selected technology companies highlights a consistent underestimation of closing prices and a tendency of the models to smooth over market volatility, pointing to a critical need for improved feature engineering and model refinement.
The paper outlines the potential of integrating a wider array of indicators to enhance model accuracy and discusses strategies to mitigate overfitting, a key issue observed through higher test RMSE values compared to training RMSE. The findings provide insights into the promising domain of machine learning in financial investment strategies and suggest directions for future research, including regularisation techniques, cross-validation, and ensemble methods, to bolster the predictive ability of machine learning models for informed stock investing decisions.
I received a High Distinction for my thesis (90/100). Feel free to download the report and have a read (it is lengthy but interesting! 🤠).