Optimal FALQON for Quantum Approximate Optimization via Layer-wise Parameter Tuning
Abstract
Feedback-based adaptive quantum optimization (FALQON) is a promising approach for combinatorial optimization on noisy intermediate-scale quantum devices, but prior studies report that standard fixed-parameter schedules can require relatively deep circuits to reach strong solution quality. We study Optimal FALQON, a layer-wise formulation that treats the time-step δk and scaling parameter wk as classical decision variables optimized per layer. The benchmark uses all 94 non-isomorphic 3-regular graphs with 12 vertices under 8192-shot simulation and compares standard FALQON, Optimal FALQON, and QAOA/QAOA-MA Powell runs warm-started from FALQON-derived schedules. Across approximation ratio, success probability, evaluation-normalized efficiency, and layer-normalized efficiency, Optimal FALQON substantially improves on standard FALQON, and Optimal-FALQON warm starts substantially improve on standard-FALQON warm starts.
Related
- Parameter-Dependent Noise Resilience in the Quantum Approximate Optimization Algorithm
- Binary Optimization with Complex Constraints via Quantum Approximate Multi-Objective Optimization
- Optimal Approximate Quantum Conversion from a Taut-String Principle in Majorization-Based Resource Theories
- Depth analysis of the Quantum Approximate Optimization Algorithm with a Grover mixer
- Quantum thermalization achieves optimal approximate quantum error correction
BibTeX
@misc{mancini2026optimal,
title = {{Optimal FALQON for Quantum Approximate Optimization via Layer-wise Parameter Tuning}},
author = {Michael Mancini and Shabnam Sodagari},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2605.08332},
eprint = {2605.08332},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.08332},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Optimal FALQON for Quantum Approximate Optimization via Layer-wise Parameter Tuning
AU - Michael Mancini
AU - Shabnam Sodagari
PY - 2026
DA - 2026/11/03
PB - arXiv
DO - 10.48550/arxiv.2605.08332
UR - https://arxiv.org/abs/2605.08332
AB - Feedback-based adaptive quantum optimization (FALQON) is a promising approach for combinatorial optimization on noisy intermediate-scale quantum devices, but prior studies report that standard fixed-parameter schedules can require relatively deep circuits to reach strong solution quality. We study Optimal FALQON, a layer-wise formulation that treats the time-step δ_k and scaling parameter w_k as classical decision variables optimized per layer. The benchmark uses all 94 non-isomorphic 3-regular graphs with 12 vertices under 8192-shot simulation and compares standard FALQON, Optimal FALQON, and QAOA/QAOA-MA Powell runs warm-started from FALQON-derived schedules. Across approximation ratio, success probability, evaluation-normalized efficiency, and layer-normalized efficiency, Optimal FALQON substantially improves on standard FALQON, and Optimal-FALQON warm starts substantially improve on standard-FALQON warm starts.
ER -