Time-domain-multiplexed Gaussian Boson sampling for graph problems in finance

Gaussian Boson sampling (GBS) has emerged as a promising platform for demonstrating quantum computational advantage. Here, we apply GBS to address classically intractable graph problems in financial security. We demonstrate algorithms for: (1) Community structure analysis for market anomaly screening; (2) Synthetic identity fraud detection via hierarchical dense subgraph identification; (3) Collusive stock manipulation detection through clique enumeration; and (4) Anti-money laundering through isomorphic subgraph matching. Notably, we design non-universal time-domain-multiplexed photonic quantum circuits to effectively map these GBS-based algorithms with a quadratic reduction in hardware complexity, achieving over 99% fidelity in fitting the target unitary matrix. This spatial-temporal hybrid architecture and a hardware-algorithm co-design framework establish a promising approach to complex financial graph analytics. Our work bridges the fields of quantum computing and finance, highlighting GBS as a near-term solution to critical industry challenges.