Keywords: Weights, Calibration, Census, Quantum computing, Quantum annealing, Global optimizer
The USDA's National Agricultural Statistics Service (NASS) conducts the U.S. Census of Agriculture every five years. The Census provides information on the characteristics of farms and the people who operate them. NASS recently developed an integer calibration (INCA) algorithm that produces integer weights required for publishing consistent Census estimates across several layers of aggregation. In INCA, a set of initial weights are first rounded to integers. Then the algorithm calibrates the rounded weights to match known population totals. However, INCA does not guarantee that the final calibrated weights are a global solution since the algorithm is based on a discrete local minimizer. This approach can be improved by quantum computing techniques. In fact, a quantum version of INCA (QUINCA) based on quantum annealing can exploit quantum tunneling properties of existing quantum algorithms to solve the calibration problem. The proposed algorithm is designed to discover global optimal solutions by performing the necessary computations with a limited amount of qubits.