Bandit Algorithm Driven by a Classical Random Walk and a Quantum Walk
arXiv:2304.10118 · doi:10.3390/e25060843
Abstract
Quantum walks (QWs) have a property that classical random walks (RWs) do not possess -- the coexistence of linear spreading and localization -- and this property is utilized to implement various kinds of applications. This paper proposes RW- and QW-based algorithms for multi-armed-bandit (MAB) problems. We show that, under some settings, the QW-based model realizes higher performance than the corresponding RW-based one by associating the two operations that make MAB problems difficult -- exploration and exploitation -- with these two behaviors of QWs.
arXiv comment21 pages, 11 figures
References in corpus (22)
- Quantum random walks - an introductory overview
- Universal computation by quantum walk
- Quantum walks: a comprehensive review
- From quantum cellular automata to quantum lattice gases
- Universal computation by multi-particle quantum walk
- Bulk--Boundary Correspondence for Chiral Symmetric Quantum Walks
- Decoherence in quantum walks - a review
- One-Dimensional Three-State Quantum Walk
- On the absence of homogeneous scalar unitary cellular automata
- Unveiling hidden topological phases of a one-dimensional Hadamard quantum walk
- The Staggered Quantum Walk Model
- Quantum recurrence of a subspace and operator-valued Schur functions
- Quantum Random Walks do not need a Coin Toss
- Quantum key distribution with quantum walks
- Inhomogeneous Quantum Walks
- Asymptotic behavior of quantum walks on the line
- Singular continuous Cantor spectrum for magnetic quantum walks
- One-Dimensional Quantum Walks with a Position-Dependent Coin
- General condition of quantum teleportation by one-dimensional quantum walks
- Lively quantum walks on cycles
- Localization of space-inhomogeneous three-state quantum walks
- New theory of diffusive and coherent nature of optical wave via a quantum walk