Human-AI Collaboration for Explainable Decision-Making in Financial Fraud Detection

Thesis Abstract

Abstract

In response to the growing complexity of financial fraud landscapes, this research addresses the critical need for human-AI collaboration in decision-making for fraud detection. The study investigates the evolution of AI in financial fraud detection, the pivotal role of explainability in machine learning, and the ethical considerations inherent in financial AI systems. The literature review provides insights into previous approaches and challenges in creating explainable models. The core of the research involves the development and evaluation of a collaborative system, uniting human expertise with machine learning capabilities for enhanced fraud detection accuracy. The implementation and evaluation phases encompass integration with financial institutions, performance metrics, user experience, ethical considerations, and compliance with regulatory standards. The outcomes contribute to the ongoing discourse on designing effective and ethical human-AI partnerships in the complex realm of financial fraud detection.