Entanglement as a Resource for Correlation: A Quantum Circuit Born Machine for System-Wide Flight-Delay Scenario Generation from Real BTS Data

Network delays are correlated, so treating hub airports as independent can severely understate simultaneous delay events. This study formulates system-wide flight-delay scenario generation with a pure-state quantum circuit Born machine (QCBM), in which each qubit represents one hub and each computational-basis measurement produces one joint binary delay scenario. Within this unitary pure-state model, a non-entangling circuit yields a factorized distribution; entangling operations therefore provide the resource required to represent cross-hub dependence. Using 24 months of U.S. Bureau of Transportation Statistics data, the five-hub exact-probability finite-difference MMD simulation reduces correlation-matrix error from 0.380 for the independent model to 0.028 ± 0.010. A finite-shot SPSA simulation using 8192 shots per circuit evaluation reaches 0.060 ± 0.029 under ideal sampling and 0.066 ± 0.026 after 0.5% noise-injected training with noiseless post-training evaluation. The independent model underestimates the observed system-wide probability by 6.6-fold for five hubs and 116-fold for 10 hubs. However, second-order classical point estimates lie within the observed five-hub bootstrap interval, and classical models evaluated over the complete 32–1024-state spaces fit efficiently. No quantum advantage or physical-hardware demonstration is claimed.