skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Quantum algorithms for Gibbs sampling and hitting-time estimation

Abstract

In this paper, we present quantum algorithms for solving two problems regarding stochastic processes. The first algorithm prepares the thermal Gibbs state of a quantum system and runs in time almost linear in √Nβ/Ζ and polynomial in log(1/ϵ), where N is the Hilbert space dimension, β is the inverse temperature, Ζ is the partition function, and ϵ is the desired precision of the output state. Our quantum algorithm exponentially improves the dependence on 1/ϵ and quadratically improves the dependence on β of known quantum algorithms for this problem. The second algorithm estimates the hitting time of a Markov chain. For a sparse stochastic matrix Ρ, it runs in time almost linear in 1/(ϵΔ 3/2), where ϵ is the absolute precision in the estimation and Δ is a parameter determined by Ρ, and whose inverse is an upper bound of the hitting time. Our quantum algorithm quadratically improves the dependence on 1/ϵ and 1/Δ of the analog classical algorithm for hitting-time estimation. Finally, both algorithms use tools recently developed in the context of Hamiltonian simulation, spectral gap amplification, and solving linear systems of equations.

Authors:
 [1]; ORCiD logo [2]
  1. Univ. of New Mexico, Albuquerque, NM (United States). Center for Quantum Information and Control; New Mexico Consortium, Los Alamos, NM (United States)
  2. Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Publication Date:
Research Org.:
Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Sponsoring Org.:
USDOE Laboratory Directed Research and Development (LDRD) Program
Contributing Org.:
New Mexico Consortium, Los Alamos, NM (United States)
OSTI Identifier:
1360697
Report Number(s):
LA-UR-16-21218
Journal ID: ISSN 1533-7146
Grant/Contract Number:
AC52-06NA25396
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
Quantum Information & Computation
Additional Journal Information:
Journal Volume: 17; Journal Issue: 1-2; Journal ID: ISSN 1533-7146
Publisher:
Rinton Press
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING; Computer Science; Information Science; Quantum algorithms

Citation Formats

Chowdhury, Anirban Narayan, and Somma, Rolando D. Quantum algorithms for Gibbs sampling and hitting-time estimation. United States: N. p., 2017. Web. doi:10.26421/QIC17.1-2.
Chowdhury, Anirban Narayan, & Somma, Rolando D. Quantum algorithms for Gibbs sampling and hitting-time estimation. United States. doi:10.26421/QIC17.1-2.
Chowdhury, Anirban Narayan, and Somma, Rolando D. Wed . "Quantum algorithms for Gibbs sampling and hitting-time estimation". United States. doi:10.26421/QIC17.1-2. https://www.osti.gov/servlets/purl/1360697.
@article{osti_1360697,
title = {Quantum algorithms for Gibbs sampling and hitting-time estimation},
author = {Chowdhury, Anirban Narayan and Somma, Rolando D.},
abstractNote = {In this paper, we present quantum algorithms for solving two problems regarding stochastic processes. The first algorithm prepares the thermal Gibbs state of a quantum system and runs in time almost linear in √Nβ/Ζ and polynomial in log(1/ϵ), where N is the Hilbert space dimension, β is the inverse temperature, Ζ is the partition function, and ϵ is the desired precision of the output state. Our quantum algorithm exponentially improves the dependence on 1/ϵ and quadratically improves the dependence on β of known quantum algorithms for this problem. The second algorithm estimates the hitting time of a Markov chain. For a sparse stochastic matrix Ρ, it runs in time almost linear in 1/(ϵΔ3/2), where ϵ is the absolute precision in the estimation and Δ is a parameter determined by Ρ, and whose inverse is an upper bound of the hitting time. Our quantum algorithm quadratically improves the dependence on 1/ϵ and 1/Δ of the analog classical algorithm for hitting-time estimation. Finally, both algorithms use tools recently developed in the context of Hamiltonian simulation, spectral gap amplification, and solving linear systems of equations.},
doi = {10.26421/QIC17.1-2},
journal = {Quantum Information & Computation},
number = 1-2,
volume = 17,
place = {United States},
year = {Wed Feb 01 00:00:00 EST 2017},
month = {Wed Feb 01 00:00:00 EST 2017}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Save / Share:
  • Hitting times for discrete quantum walks on graphs give an average time before the walk reaches an ending condition. To be analogous to the hitting time for a classical walk, the quantum hitting time must involve repeated measurements as well as unitary evolution. We derive an expression for hitting time using superoperators, and numerically evaluate it for the discrete walk on the hypercube. The values found are compared to other analogs of hitting time suggested in earlier work. The dependence of hitting times on the type of unitary 'coin' is examined, and we give an example of an initial statemore » and coin which gives an infinite hitting time for a quantum walk. Such infinite hitting times require destructive interference, and are not observed classically. Finally, we look at distortions of the hypercube, and observe that a loss of symmetry in the hypercube increases the hitting time. Symmetry seems to play an important role in both dramatic speed-ups and slow-downs of quantum walks.« less
  • We present an adiabatic quantum algorithm for the abstract problem of searching marked vertices in a graph, or spatial search. Given a random walk (or Markov chain) P on a graph with a set of unknown marked vertices, one can define a related absorbing walk P{sup '} where outgoing transitions from marked vertices are replaced by self-loops. We build a Hamiltonian H(s) from the interpolated Markov chain P(s)=(1-s)P+sP{sup '} and use it in an adiabatic quantum algorithm to drive an initial superposition over all vertices to a superposition over marked vertices. The adiabatic condition implies that, for any reversible Markovmore » chain and any set of marked vertices, the running time of the adiabatic algorithm is given by the square root of the classical hitting time. This algorithm therefore demonstrates a novel connection between the adiabatic condition and the classical notion of hitting time of a random walk. It also significantly extends the scope of previous quantum algorithms for this problem, which could only obtain a full quadratic speedup for state-transitive reversible Markov chains with a unique marked vertex.« less
  • We present a quantum algorithm to prepare the thermal Gibbs state of interacting quantum systems. This algorithm sets a universal upper bound D{sup {alpha}} on the thermalization time of a quantum system, where D is the system's Hilbert space dimension and {alpha}<=(1/2) is proportional to the Helmholtz free energy density. We also derive an algorithm to evaluate the partition function of a quantum system in a time proportional to the system's thermalization time and inversely proportional to the targeted accuracy squared.
  • Many of the commonly-used groundwater sampling techniques and procedures have resulted from methods developed for water supply investigations. These methods have persisted, even though the monitoring goals may have changed from water supply development to contaminant source and plume delineation. Unfortunately, the use of these methods can result in an incorrect understanding of the contaminant concentrations and system geochemistry, often overestimating or underestimating the true values or the extent of a plume. It was discovered, during investigations of potential contaminant transport by mobile colloidal particles, that traditional purging and sampling were disrupting the aquifer and sand pack materials. The disruptionmore » was caused by the stress of high water velocities and/or water surging in the wells due to sampling with bailers and high-speed submersible pumps. This left no conclusive way to determine whether collected particulates in the samples were naturally occurring or due to the disturbance caused by the sampling device/approach. Such induced stresses can cause grain flow within the sand pack and/or exceed the cohesive forces of aquifer mineral cementation, resulting in excessive and artificial sample turbidity. This increased turbidity can radically alter analytical results for water samples, causing spurious in creases in analyzed metal concentrations.« less
  • Hitting times are the average time it takes a walk to reach a given final vertex from a given starting vertex. The hitting time for a classical random walk on a connected graph will always be finite. We show that, by contrast, quantum walks can have infinite hitting times for some initial states. We seek criteria to determine if a given walk on a graph will have infinite hitting times, and find a sufficient condition, which for discrete time quantum walks is that the degeneracy of the evolution operator be greater than the degree of the graph. The set ofmore » initial states which give an infinite hitting time form a subspace. The phenomenon of infinite hitting times is in general a consequence of the symmetry of the graph and its automorphism group. Using the irreducible representations of the automorphism group, we derive conditions such that quantum walks defined on this graph must have infinite hitting times for some initial states. In the case of the discrete walk, if this condition is satisfied the walk will have infinite hitting times for any choice of a coin operator, and we give a class of graphs with infinite hitting times for any choice of coin. Hitting times are not very well defined for continuous time quantum walks, but we show that the idea of infinite hitting-time walks naturally extends to the continuous time case as well.« less