Litcius/Paper detail

A Tutorial on Quantum Graph Recurrent Neural Network (QGRNN)

Jaeho Choi, Seunghyeok Oh, Joongheon Kim

202121 citationsDOI

Abstract

Over the past decades, various neural networks have been proposed with the rapid development of the machine learning field. In particular, graph neural networks using feature-vectors assigned to nodes and edges have been attracting attention in various fields. The usefulness of graph neural networks also affected the field of quantum computing, which led to the birth of quantum graph neural networks composed of parameterized quantum circuits. The quantum graph neural networks have many possibilities as applications from the simulation perspective of quantum dynamics. Among the application models of various quantum graph neural networks, the quantum graph recurrent neural network (QGRNN) is proven to be effective in training the Ising model Hamiltonian. Thus, this paper introduces the concepts of the Ising model, variational quantum eigensolver (VQE) for preparing quantum data, and QGRNN from a software engineer's point of view.

Topics & Concepts

Computer scienceQuantum machine learningArtificial neural networkQuantum computerQuantumIsing modelTheoretical computer scienceGraphArtificial intelligenceStatistical physicsPhysicsQuantum mechanicsQuantum Computing Algorithms and ArchitectureQuantum many-body systemsQuantum and electron transport phenomena