Hybrid Quantum-Classical Clustering for Preparing a Prior Distribution of Eigenspectrum

TL;DR

This paper introduces a hybrid quantum-classical algorithm for preparing a prior eigenspectrum distribution of Hamiltonians, aiding in quantum eigenvalue problems with applications to physical systems.

Contribution

It proposes a novel three-step method combining Hamiltonian transformation, parameter representation, and classical clustering to approximate eigenspectra.

Findings

Effective in 1D Heisenberg system

Applicable to LiH molecule

Shows scalability and resource efficiency

Abstract

Determining the energy gap in a quantum many-body system is critical to understanding its behavior and is important in quantum chemistry and condensed matter physics. The challenge of determining the energy gap requires identifying both the excited and ground states of a system. In this work, we consider preparing the prior distribution and circuits for the eigenspectrum of time-independent Hamiltonians, which can benefit both classical and quantum algorithms for solving eigenvalue problems. The proposed algorithm unfolds in three strategic steps: Hamiltonian transformation, parameter representation, and classical clustering. These steps are underpinned by two key insights: the use of quantum circuits to approximate the ground state of transformed Hamiltonians and the analysis of parameter representation to distinguish between eigenvectors. The algorithm is showcased through applications to the 1D Heisenberg system and the LiH molecular system, highlighting its potential for both near-term quantum devices and fault-tolerant quantum devices. The paper also explores the scalability of the method and its performance across various settings, setting the stage for more resource-efficient quantum computations that are both accurate and fast. The findings presented here mark a new insight into hybrid algorithms, offering a pathway to overcoming current computational challenges.