A quantum-inspired classical algorithm for recommendation systems

DOI10.1145/3313276.3316310MaRDI QIDQ5212763zbMATH OpenOpenAlexWikidataFDO

Authors Ewin Tang

Publication date 30 January 2020

Published in Proceedings of the 51st Annual ACM SIGACT Symposium on Theory of Computing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1807.04271

zbMATH Keywords

exponential speeduplow-rank approximationquantum machine learningrecommender systemssampling

Mathematics Subject Classification ID

Data structures (68P05) Quantum algorithms and complexity in the theory of computing (68Q12) Learning and adaptive systems in artificial intelligence (68T05) Theory of languages and software systems (knowledge-based systems, expert systems, etc.) for artificial intelligence (68T35) Analysis of algorithms (68W40)

Abstract: We give a classical analogue to Kerenidis and Prakash's quantum recommendation system, previously believed to be one of the strongest candidates for provably exponential speedups in quantum machine learning. Our main result is an algorithm that, given an $mimesn$ matrix in a data structure supporting certain $el{l}^2$-norm sampling operations, outputs an $el{l}^2$-norm sample from a rank-$k$ approximation of that matrix in time $O(extpoly(k)log(mn))$, only polynomially slower than the quantum algorithm. As a consequence, Kerenidis and Prakash's algorithm does not in fact give an exponential speedup over classical algorithms. Further, under strong input assumptions, the classical recommendation system resulting from our algorithm produces recommendations exponentially faster than previous classical systems, which run in time linear in $m$ and $n$. The main insight of this work is the use of simple routines to manipulate $el{l}^2$-norm sampling distributions, which play the role of quantum superpositions in the classical setting. This correspondence indicates a potentially fruitful framework for formally comparing quantum machine learning algorithms to classical machine learning algorithms.

Recommendations

• Quantum recommendation systems
• Sampling-based sublinear low-rank matrix arithmetic framework for dequantizing quantum machine learning
• Quantum machine learning: a classical perspective
• Quantum context-aware recommendation systems based on tensor singular value decomposition
• A survey on HHL algorithm: from theory to application in quantum machine learning

Cited in

(28)

• Quantum learning of concentrated Boolean functions
• A survey on HHL algorithm: from theory to application in quantum machine learning
• Quantum-enhanced least-square support vector machine: simplified quantum algorithm and sparse solutions
• Sample-size-reduction of quantum states for the noisy linear problem
• Quantum context-aware recommendation systems based on tensor singular value decomposition
• The road to quantum computational supremacy
• Quantum recommendation systems
• scientific article; zbMATH DE number 7559394 (Why is no real title available?)
• On solving linear systems in sublinear time
• Quantum SDP solvers: large speed-ups, optimality, and applications to quantum learning
• Sampling-based sublinear low-rank matrix arithmetic framework for dequantizing quantum machine learning
• Sampling-based Sublinear Low-rank Matrix Arithmetic Framework for Dequantizing Quantum Machine Learning
• scientific article; zbMATH DE number 7765405 (Why is no real title available?)
• Quantum-inspired canonical correlation analysis for exponentially large dimensional data
• The quantum detection of projectors in finite-dimensional algebras and holography
• Quantum algorithm for computing distances between subspaces
• Quantum-inspired algorithm for truncated total least squares solution
• Multi-class classification based on quantum state discrimination
• Quantum speedups for linear programming via interior point methods
• Guidable local Hamiltonian problems with implications to heuristic ansatz state preparation and the quantum PCP conjecture
• Exponential speedup of quantum algorithms for the pathfinding problem
• Robust dequantization of the quantum singular value transformation and quantum machine learning algorithms
• Quantum communication complexity of linear regression
• Multivariate trace estimation using quantum state space linear algebra
• Quantum advantage for multi-option portfolio pricing and valuation adjustments
• Quantum time-space tradeoffs for matrix problems
• Quantum-inspired methods for finite-element discretizations of the high-dimensional Poisson equation
• Classical algorithms for constant approximation of the ground state energy of local Hamiltonians

This page was built for publication: A quantum-inspired classical algorithm for recommendation systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5212763)