Abstract:We propose fast and practical quantum-inspired classical algorithms for solving linear systems. Specifically, given sampling and query access to a matrix $A\in\mathbb{R}^{m\times n}$ and a vector $b\in\mathbb{R}^m$, we propose classical algorithms that produce a data structure for the solution $x\in\mathbb{R}^{n}$ of the linear system $Ax=b$ with the ability to sample and query its entries. The resulting $x$ satisfies $\|x-A^{+}b\|\leq\epsilon\|A^{+}b\|$, where $\|\cdot\|$ is the spectral norm and $A^+$ is the Moore-Penrose inverse of $A$. Our algorithm has time complexity $\widetilde{O}(\kappa_F^4/\kappa\epsilon^2)$ in the general case, where $\kappa_{F} =\|A\|_F\|A^+\|$ and $\kappa=\|A\|\|A^+\|$ are condition numbers. Compared to the prior state-of-the-art result [Shao and Montanaro, arXiv:2103.10309v2], our algorithm achieves a polynomial speedup in condition numbers. When $A$ is $s$-sparse, our algorithm has complexity $\widetilde{O}(s \kappa\log(1/\epsilon))$, matching the quantum lower bound for solving linear systems in $\kappa$ and $1/\epsilon$ up to poly-logarithmic factors [Harrow and Kothari]. When $A$ is $s$-sparse and symmetric positive-definite, our algorithm has complexity $\widetilde{O}(s\sqrt{\kappa}\log(1/\epsilon))$. Technically, our main contribution is the application of the heavy ball momentum method to quantum-inspired classical algorithms for solving linear systems, where we propose two new methods with speedups: quantum-inspired Kaczmarz method with momentum and quantum-inspired coordinate descent method with momentum. Their analysis exploits careful decomposition of the momentum transition matrix and the application of novel spectral norm concentration bounds for independent random matrices. Finally, we also conduct numerical experiments for our algorithms on both synthetic and real-world datasets, and the experimental results support our theoretical claims.

Efficient quantum algorithms for solving quantum linear system problems

Fast and Practical Quantum-Inspired Classical Algorithms for Solving Linear Systems.

An Efficient Quantum Algorithm for Linear System Problem in Tensor Format

Quantum linear system algorithm with optimal queries to initial state preparation

Efficient quantum linear solver algorithm with detailed running costs

Quantum eigenvalue processing

Quantum multi-row iteration algorithm for linear systems with non-square coefficient matrices

Quantum Algorithm For Solving Nonlinear Algebraic Equations

A shortcut to an optimal quantum linear system solver

Quantum algorithms for matrix operations and linear systems of equations

Quantum Algorithm for Solving a Quadratic Nonlinear System of Equations

A quantum algorithm for solving systems of nonlinear algebraic equations

Efficient State Read-out for Quantum Machine Learning Algorithms

Quantum Linear System Solver Based on Time-optimal Adiabatic Quantum Computing and Quantum Approximate Optimization Algorithm

Towards efficient variational quantum algorithm for solving linear equations

Qubit-Efficient Randomized Quantum Algorithms for Linear Algebra

The Algorithm for Solving Quantum Linear Systems of Equations With Coherent Superposition and Its Extended Applications

Quantum Discrete Adiabatic Linear Solver based on Block Encoding and Eigenvalue Separator

Experimental Quantum Computing to Solve Systems of Linear Equations

A near-term quantum algorithm for solving linear systems of equations based on the Woodbury identity