Abstract Hardware Error Outputs of Quantum Computing Not Reliable because of Hardware Calibration Drift, Measurement Noise, Gate Errors on NISQ Systems. Even though MEMs, ZNEs, and CDRs slightly increase output accuracy, they operate under the assumption of static noise, therefore they do not account for changing output of the NISQ devices in real-time. The limitations of MEMs, ZNEs, and CDRs can be overcome by using the Hybrid Adaptive Error Mitigation (HAEM) method proposed in this article that has the capability to combine baseline measurement-error mitigation with Real-time Calibration Circuits and Lightweight-Machine-learning algorithms for adaptive error-correcting and increasing the overall output integrity of quantum devices affected by fluctuations in noise. The application of the HAEM method will utilize Qiskit Runtime to develop, implement, and validate the Hybrid Adaptive Error Mitigation method through testing using controlled noisy simulated environments based on the characteristics of current advanced NISQ quantum computers. The results of the tests demonstrate the HAEM architecture provides a higher level of accuracy and fidelity than static and non-mitigated executions ("mitigated execution vs. HAEM execution"), providing an increase of up to thirty-four percent in fidelity under high-noise conditions with minimal computational overhead. The results demonstrate that the use of adaptive and data-driven mitigation techniques, like those used in the Hybrid Adaptive Error Mitigation method, are an efficient way to increase the accuracy of quantum computations.
Karthick et al. (Wed,) studied this question.