A new approach to rating scale definition with quantum-inspired optimization
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
In finance, assessing the creditworthiness of loan applicants requires lenders to cluster borrowers using rating scales. Financial institutions must define the scales in compliance with strict institutional constraints, resulting in solving a complex combinatorial constrained optimization problem. This contribution studies how to solve this problem using a Quadratic Unconstrained Binary Optimization (QUBO) model, a formulation suitable for quantum hardware. We validate this approach by testing the proposed formulation with classical heuristics. We then benchmark the results against a brute-force method to demonstrate consistent solution quality and highlight the framework's suitability for more complex scenarios.
Fields of Study & Subjects
Metaheuristic Optimization Algorithms Research
Quantum Computing Algorithms and Architecture
Advanced Optimization Algorithms Research
Publication details
- Journal / source
- arXiv (Cornell University)
- Publication date
- 27 March 2026
- Language
- EN
- Document type
- Preprint