Investigate the Role of AI and ML in Predicting and Preventing Financial Fraud
Kikani, Parth (2025)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-202503264993
https://urn.fi/URN:NBN:fi:amk-202503264993
Tiivistelmä
The objective of this thesis was to investigate the role of Artificial Intelligence (AI) and Machine Learning (ML) in predicting and preventing financial fraud. In contrast to traditional approaches like rule-based systems and manual reviews, it looks at how AI-based techniques can improve fraud detection in the financial sector. The study intends to investigate AI's potential in enhancing detection accuracy and efficiency considering the rise in financial crimes like identity theft, Ponzi schemes, and phishing.
Based on a thorough review of the existing literature, the study investigates deep learning methods as well as supervised, unsupervised, and reinforcement learning machine learning techniques. To demonstrate how AI can automate fraud detection and lower false positives by spotting odd patterns and anomalies, real-world applications like anomaly detection and predictive analytics are covered.
The results demonstrate how AI can improve real-time analysis and leverage big datasets to improve fraud detection procedures. However, the study also discusses issues like limitations in data quality, scalability, and ethical concerns. To increase the efficacy and scalability of AI-based fraud detection systems, recommendations include enhancing the interpretability of AI models, addressing privacy issues, and optimizing integration with current financial infrastructures.
Based on a thorough review of the existing literature, the study investigates deep learning methods as well as supervised, unsupervised, and reinforcement learning machine learning techniques. To demonstrate how AI can automate fraud detection and lower false positives by spotting odd patterns and anomalies, real-world applications like anomaly detection and predictive analytics are covered.
The results demonstrate how AI can improve real-time analysis and leverage big datasets to improve fraud detection procedures. However, the study also discusses issues like limitations in data quality, scalability, and ethical concerns. To increase the efficacy and scalability of AI-based fraud detection systems, recommendations include enhancing the interpretability of AI models, addressing privacy issues, and optimizing integration with current financial infrastructures.