Speaker
Description
In the field of finance, lenders must estimate the creditworthiness of loan applicants, which leads them to devise strategies for classifying borrowers. Among these strategies, banking groups identify the definition of a rating scale as a classification method regulated by national and international constraints. The definition of a rating scale can be described as a combinatorial optimization problem.
The mathematical formulation known as Quadratic Unconstrained Binary Optimization (QUBO) can address a wide variety of significant combinatorial optimization problems, including rating scale definition. It has been proven that this kind of formulation can benefit from quantum computing due to its equivalence to a quantum system described by an Ising Hamiltonian.
In the context of the ICSC – Centro Nazionale di Ricerca in HPC, Big Data and Quantum Computing, a collaboration project between INFN and Intesa Sanpaolo Bank aims to build a QUBO formulation to define a rating scale for classifying borrowers. We present the final results of this project, starting from the definition of the financial constraints up to its implementation and testing using different solver simulators. Furthermore, we compare the results of a classical implementation with the performance obtained using a quantum simulator.
| Sessions | Quantum Simulation |
|---|---|
| Invited | No |