Constructing Compact ADAPT Unitary Coupled-Cluster Ansatz with Parameter-Based Criterion

TL;DR

This paper introduces Param-ADAPT-VQE, an improved adaptive quantum algorithm that reduces measurement costs and eliminates redundant operators, enhancing scalability and accuracy for molecular ground state energy calculations.

Contribution

It proposes a parameter-based operator selection criterion and integrates a sub-Hamiltonian technique with hot-start optimization, significantly improving ADAPT-VQE's efficiency.

Findings

Outperforms original ADAPT-VQE in accuracy and measurement costs

Reduces redundant excitation operators effectively

Compatible with various ADAPT-VQE modifications

Abstract

The adaptive derivative-assembled pseudo-trotter variational quantum eigensolver (ADAPT-VQE) is a promising hybrid quantum-classical algorithm for molecular ground state energy calculation, yet its practical scalability is hampered by redundant excitation operators and excessive measurement costs. To address these challenges, we propose Param-ADAPT-VQE, a novel improved algorithm that selects excitation operators based on a parameter-based criterion instead of the traditional gradient-based metric. This strategy effectively eludes redundant operators. We further develop a sub-Hamiltonian technique and integrate a hot-start VQE optimization strategy, achieving a significant reduction in measurement costs. Numerical experiments on typical molecular systems demonstrate that Param-ADAPT-VQE outperforms the original ADAPT-VQE in computational accuracy, ansatz size, and measurement costs. Furthermore, our scheme retains the fundamental framework of ADAPT-VQE and is thus fully compatible with its various modified versions, enabling further performance improvements in specific aspects. This work presents an efficient and scalable enhancement to ADAPT-VQE, mitigating the core obstacles that impede its practical implementation in the field of molecular quantum chemistry.