Sadasivan, Praveen (2026) Application of artificial intelligence models for prediction of Australian share market index for financial sector. PhD thesis, Victoria University.

Abstract

Stock market prediction is a complex task due to the highly nonlinear and dynamic nature of financial time-series data. Recent advances in Artificial Intelligence and Machine Learning have enabled data-driven approaches capable of capturing complex market patterns without relying on rigid statistical assumptions. While extensive studies have been conducted on major international stock exchanges such as those in the United States, Europe, and Asia, comparatively fewer studies have focused on the effectiveness of these techniques on Australian financial market. This study investigated and compared the performance of Artificial Neural Networks, Long Short-Term Memory networks, and Deep Learning models for stock price prediction using historical stock market data. The research contributes to the existing body of knowledge by addressing the gap in literature related to Machine Learning applications in the Australian stock market. The study adopted a structured, comparative modelling framework incorporating both traditional statistical methods and advanced machine learning approaches. Four independent variables were considered for the study, namely the NASDAQ Index, crude oil prices, foreign exchange rates, and the London Stock Exchange Index, while the dependent variable was the Australian financial sector stock market index. The models were trained and evaluated across multiple EPOCHS using standard performance metrics to assess convergence behaviour and predictive accuracy. The experimental results indicated that Deep Learning and LSTM models consistently outperform traditional ANN models, particularly in modelling time-based dependencies and reducing prediction error. The main contents and contributions of this thesis was as follows. 1. This thesis explored the predictive effectiveness of machine learning models in the context of Australian share market index forecasting, while explicitly considering the trade-offs between model complexity, computational efficiency, ability to interpret, and extent of predictive accuracy. 2. Chapter 1 provided the background and motivation for this research and detailed the research problem, hypothesis, thesis contribution and the methodology. 3. Chapter 2 presented a detailed literature review. 4. Chapter 3 employed statistical methods for share market prediction in ASX context. 5. Chapter 4 examined AI models-based predictions by optimising model parameters like hidden layers and EPCOHS. 6. Chapter 5 analysed and compared AI models for predictive efficiency in ASX context and chose the most suited AI Model. 7. Chapter 6 provided a conclusion of this study and presented areas for future research. The selected Deep Learning model demonstrated a balanced design that achieves strong predictive performance without excessive training complexity or reliance on extensive feature engineering. The findings provide valuable insights into the suitability of advanced AI-based forecasting models for Australian financial markets and offer practical implications for investors and researchers. This research makes a substantive contribution to the literature by providing evidence on the applicability and effectiveness of advanced predictive analytics in the Australian stock market.

Search Google Scholar

Repository staff login