Enhancing Risk Management in Financial Institutions through Machine Learning and Natural Language Processing: A Critical Evaluation of AI Techniques

Elias, Anjaly (2025) Enhancing Risk Management in Financial Institutions through Machine Learning and Natural Language Processing: A Critical Evaluation of AI Techniques. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (735kB) | Preview
Preview	PDF (Configuration Manual) Download (313kB) | Preview

Abstract

The financial services industry is undergoing a significant transformation driven by the integration of artificial intelligence. Effective risk management, a cornerstone of institutional stability, stands to benefit immensely from these technological advancements. This thesis explores the application of machine learning (ML) and natural language processing (NLP) to enhance credit risk assessment, a critical component of financial risk management. The whole Lending Club loan data set will be used in establishing a reliable baseline for modelling, following a robust data pre-processing, balancing, and feature engineering process. This research will evaluate the performance of three distinct AI methods- an ensemble method from the historical world (Random Forest), a probabilistic classifier (Gaussian Naive Bayes), and a deep learning model (Long Short Term Memory network). All of those models make use of a balanced dataset that consists of 100,000 loan applications. The approaches are rigorously compared by performance measurement of accuracy, precision, recall, F1 score, and AUC. The results show that LSTM models achieved the highest predictive accuracy of 73.55%, improving the tuned random forest performance of 73.14% and outpacing significantly naive Bayes models with 70.23%. This shows that deep learning architectures have the potential to capture complex, non linear patterns in financial data that can be missed by traditional models. The last phase of research involves design creation and implementation of a prototype web application demonstrating a real avenue through which advanced analytic might be deployed in a financial institution's workflow. Findings point towards the promise of AI to upgrade risk predictions models and painstaking data preparation and evaluation processes.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Del Rosal, Victor UNSPECIFIED
Subjects:	H Social Sciences > HD Industries. Land use. Labor > HD61 Risk Management Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing H Social Sciences > HG Finance > Fintech T Technology > T Technology (General) > Information Technology > Fintech Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in FinTech
Depositing User:	Ciara O'Brien
Date Deposited:	20 Aug 2026 10:40
Last Modified:	20 Aug 2026 10:40
URI:	https://norma.ncirl.ie/id/eprint/9568

Actions (login required)

View Item