Quantum Gambling: Best-Arm Strategies for Generator Selection in Adaptive Variational Algorithms

Area of Science:

• Quantum computing
• Computational chemistry
• Quantum algorithms

Background:

• Adaptive variational algorithms require extensive measurements for generator selection, limiting scalability on near-term quantum devices.
• Estimating energy gradients for large operator pools presents a significant bottleneck in current quantum simulations.

Purpose of the Study:

• To reduce the measurement overhead in adaptive variational algorithms.
• To enhance the practicality of adaptive variational algorithms for near-term quantum simulations.

Main Methods:

• Reformulated generator selection as a Best Arm Identification (BAI) problem.
• Applied the Successive Elimination (SE) algorithm for adaptive measurement allocation.
• Performed numerical experiments on molecular systems.

Main Results:

• Substantially reduced the number of required measurements.
• Preserved high ground-state energy accuracy.
• Demonstrated improved efficiency for adaptive variational algorithms.

Conclusions:

• The BAI reformulation with the SE algorithm makes adaptive variational algorithms more feasible for near-term quantum computing.
• Reduced measurement costs without compromising simulation accuracy.
• Enables simulations of larger molecular systems on current quantum hardware.