2021/10/15 by Xin Wang, Wang, Xin
Computer Science · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #FOS: Physical sciences #Neural Networks and Reservoir Computing #Quantum Computing Algorithms and Architecture #Quantum Information and Cryptography #Quantum Physics (quant-ph)
paper · pdf · doi:10.48550/arxiv.2110.08016
openalex publication_date 2021/10/15 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
Optimizing parameterized quantum circuits promises efficient use of near-term\nquantum computers to achieve the potential quantum advantage. However, there is\na notorious tradeoff between the expressibility and trainability of the\nparameter ansatz. We find that in combinatorial optimization problems, since\nthe solutions are described by bit strings, one can trade the expressiveness of\nthe ansatz for high trainability. To be specific, by focusing on the max-cut\nproblem we introduce a simple yet efficient algorithm named Quantum Qubit\nRotation Algorithm (QQRA). The quantum circuits are comprised with single-qubit\nrotation gates implementing on each qubit. The rotation angles of the gates can\nbe trained free of barren plateaus. Thus, the approximate solution of the\nmax-cut problem can be obtained with probability close to 1. To illustrate the\neffectiveness of QQRA, we compare it with the well known quantum approximate\noptimization algorithm and the classical Goemans-Williamson algorithm.\n