Quantum Machine Learning in Finance

Overview

This research explores the intersection of quantum computing and machine learning in financial applications, specifically focusing on credit default prediction. By leveraging hybrid quantum-classical models, we achieved significant improvements over traditional XGBoost baselines.

Key Findings

• • Achieved 41 basis point improvement over XGBoost baseline in credit default prediction
• • Developed hybrid quantum-classical neural network architectures
• • Collaborated with industry partners including PayPal and American Express
• • Demonstrated practical quantum advantage in specific financial use cases

Methodology

The approach combines classical machine learning preprocessing with quantum circuits for feature encoding and transformation. This hybrid architecture allows us to leverage the strengths of both paradigms while working within the constraints of current quantum hardware.

Technologies Used