Welcome to DBS eSource
DBS eSource is an online service hosting full content materials produced by Dublin Business School staff and students. It contains the full text of articles, theses, conference papers, book chapters and more. DBS eSource is an open access repository, with the aim of making all content as widely accessible as possible. Use the Browse functions on the right for an overview of relevant materials. For an advanced search click here
For more information on our institutional repository policy or the steps involved in submitting work to Esource click here
Recent Submissions
Item Loan Approval Prediction Using Machine Learning: A Data-driven Approach to Binary Classification(Dublin Business School, 2025.17.12)This research analyzes the usage of machine learning techniques for predicting loan approval accuracy and fairness. The research employs data from real world credit risk datasets and SMOTENC to tackle the class imbalance problem in building a robust binary classification model. Such methods implemented include Logistic Regression and Random Forest algorithms, and evaluated through the metrics accuracy, F1 score, and AUC ROC. The data privacy and bias mitigation are prioritized. The findings offer a replicable framework for efficient and inclusive evaluation of credit risk, and therefore contribute to innovative finance decision making.Item Evaluating the impact of ESG on firms' performance in financial sector : insights, challenges and path forward(Dublin Business School, 2025.17.12)This dissertation examines how adoption of Environmental, Social and Governance (ESG) can influence the financial and operational performance of small and medium-sized enterprises (SMEs) in the financial sectors of India and Ireland. Applying a pragmatic ideology and an explanatory sequential mixed-design method, the research used both quantitative analysis (as per SPSS) and qualitative analysis (as per NVIVO), in terms of thematic analysis. The results indicated that ESG adoption has a significant positive influence on brand reputation and operations effectiveness, which proved to be powerful determinants of financial success. However, the demand of the investors did not exhibit a statistically significant effect, indicating the difficulty that SMEs find in aligning to the demands of investors as a resource-constrained situation. The quantitative findings were supported by qualitative themes that confirmed the stakeholder power, difficulties in reporting and cultural obstacles and outlined the methods to be used to improve the situation. The research makes a theoretical contribution by generalizing Stakeholder Theory, RBV and Legitimacy Theory to the setting of SMEs and provides practical, policy and theoretical recommendations that can be used to drive healthy adoption of sustainable finance.Item The Influence of Artificial Intelligence in Risk Management Practices Among Irish Construction Industry(Dublin Business School, 2025.17.12)This dissertation investigates the adoption of Artificial Intelligence (AI) in risk management within the Irish construction industry. Grounded in the Theory of Planned Behaviour (TPB), the study examines professionals’ perceptions of AI’s usefulness, identifies barriers to adoption, and evaluates the role of organisational support. A quantitative survey of 108 industry participants was conducted, with data analysed using descriptive statistics, cross-tabulations, ANOVA, correlation, and regression. Findings reveal positive perceptions of AI’s potential to improve accuracy, efficiency, and decision-making, yet adoption is constrained by high costs, limited training, and cultural resistance. Regression results indicate that attitudes and perceived behavioural control are significant predictors of adoption intentions, while subjective norms are less influential. The study contributes to theory by validating TPB in the Irish construction context and offers practical recommendations for industry, policymakers, and technology providers to enhance AI adoption in risk management.Item Exploring the Predictive Power of Trading Volume on Stock Price Movements(Dublin Business School, 2025.17.12)This study examines whether trading volume significantly contributes to the predictive accuracy of stock price forecasting models. Trading volume is often viewed as a proxy for liquidity and sentiment, yet its independent predictive value remains unclear. Using daily S&P 500 index data from 2010 to 2025, three experimental datasets were constructed: a Volume-Only dataset, a Price-Only dataset, and a combined dataset with Principal Component Analysis (PCA) applied. Four machine learning algorithms—Random Forest, Support Vector Regression (SVR), XGBoost, and LightGBM—were tested under a leakage-free chronological split. Results show that trading volume alone had no predictive power, with models yielding negative R² values. Price-only models performed strongly, with SVR achieving R² = 0.8770, confirming the persistence of stock prices as the dominant predictor. Adding volume did not improve accuracy, and in some cases reduced performance. PCA confirmed that volume represented an independent but weak component. The results indicate that trading volume adds very little supplementary information, but historical prices remain the primary factors of forecasting accuracy.Item Forecasting and Trend Analysis of Electric Vehicle Sales Energy Consumption and Oil Displacement in the European region using machine learning(Dublin Business School, 2025.17.12)The adoption of Electric Vehicles(EVs) is rapidly increasing around the world. Europe is a major region in the world which has a lot of importance .The trend of adopting EVs is also prevalent in Europe. If forecasting related to EVs can be done in Europe then government bodies in Europe will be able to get a better understanding of the scenario related to EVs. In the study proposed here, a machine learning based model for the prediction EV sales, oil displacement, and electricity consumption by EVs is built. The dataset containing data associated with EVs was used in the study. The data was used for generating visual plots for finding patterns and trends. The Voting Regressor, LSTM, Prophet, ARIMA and SARIMA models were used in the study. The Voting Regressor contained the Random Forest(RF) and Linear Regression, and the prediction by these models were used. The models were trained using the data associated with EVs in the dataset. The results of the study showed that the models were able to successfully forecast EV sales, oil displacement, and electricity consumption. Based on the performance metrics associated with the models it was seen that the best performance in prediction was shown by Voting Regressor for the prediction of EV sales and electricity consumption. ARIMA model for the prediction of the oil displacement
Communities in DBS eSource
Select a community to browse its collections.