The Theory of Hybrid Stochastic Algorithms

  • A. D. Kennedy
Part of the NATO ASI Series book series (NSSB, volume 224)

Abstract

These lectures introduce the family of Hybrid Stochastic Algorithms for performing Monte Carlo calculations in Quantum Field Theory. After explaining the basic concepts of Monte Carlo integration we discuss the properties of Markov processes and one particularly useful example of them: the Metropolis algorithm. Building upon this framework we consider the Hybrid and Langevin algorithms from the viewpoint that they are approximate versions of the Hybrid Monte Carlo method; and thus we are led to consider Molecular Dynamics using the Leapfrog algorithm. The lectures conclude by reviewing recent progress in these areas, explaining higher-order integration schemes, the asymptotic large-volume behaviour of the various algorithms, and some simple exact results obtained by applying them to free field theory. It is attempted throughout to give simple yet correct proofs of the various results encountered.

Keywords

Sugar Manifold Autocorrelation Alan 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    John H. Halton: A retrospective and prospective survey of the Monte Carlo method. SIAM Review, 12: 1–63, 1970.MathSciNetGoogle Scholar
  2. [2]
    F. James: Monte Carlo in theory and practice. Rept. Prog. Phys., 43: 1145, 1980.ADSCrossRefGoogle Scholar
  3. [3]
    Olivier Martin and Steve W. Otto: Reducing the number of flavors in the microcanonical method. Phys. Rev., D31 (2): 435–437, January 1985.ADSGoogle Scholar
  4. [4]
    Janos Polonyi, H. W. Wyld, and John B. Kogut: Finite-temperature phase transitions in SU (3) lattice gauge theory with dynamical, light fermions. Phys. Rev. Lett., 53 (7): 644–647, August 1984.ADSCrossRefGoogle Scholar
  5. [5]
    N. A. Metropolis, M. N. Rosenbluth, A. H. Rosenbluth, E. Teller, and J. Teller: J. Chem. Phys., 21: 1087, 1953.ADSCrossRefGoogle Scholar
  6. [6]
    S. Duane: Stochastic quantization vs. the microcanonical ensemble: getting the best of both worlds. Nucl. Phys., B257 [FS14]: 652, 1985.ADSCrossRefGoogle Scholar
  7. [7]
    Simon Duane and John B. Kogut: Hybrid stochastic differential equations applied to quantum chromo dynamics. Phys. Rev. Lett., 55 (25): 2774–2777, December 1985.ADSCrossRefGoogle Scholar
  8. [8]
    Simon Duane and John B. Kogut: The theory of hybrid stochastic algorithms. Nucl. Phys., B275: 398, 1986.ADSCrossRefGoogle Scholar
  9. [9]
    Steven Gottlieb, W. Liu, D. Toussaint, R. L. Renken, and R. L. Sugar: Empirical study of the hybrid-molecular-dynamics approach to the simulation of QCD. Phys. Rev., D36 (12): 3797–3803, December 1987.ADSGoogle Scholar
  10. [10]
    H. C. Andersen: J. Chem. Phys., 72: 2384, 1980.ADSCrossRefGoogle Scholar
  11. [11]
    David J. E. Callaway and Aneesur Rahman: Microcanonical ensemble formulation of lattice gauge theory. Phys. Rev. Lett., 49 (9): 613–616, 1982.ADSCrossRefGoogle Scholar
  12. [12]
    Arunabha Guha and S.-C. Lee: Stochastic quantization of matrix and lattice gauge models. Phys. Rev., D27 (10): 2412–2422, May 1983.MathSciNetADSGoogle Scholar
  13. [13]
    J. Polonyi and H. W. Wyld: Micro canonical simulation of fermionic systems. Phys. Rev. Lett., 51:2257, 1983. Erratum: ibid. 52: 401, 1984.ADSCrossRefGoogle Scholar
  14. [14]
    J. Polonyi and H. W. Wyld: A micro canonical method for fermionic systems. Nucl. Phys., A418: 491C - 498C, 1984.ADSCrossRefGoogle Scholar
  15. [15]
    Simon Duane, A. D. Kennedy, Brian J. Pendleton, and Duncan Roweth: Hybrid Monte Carlo. Phys. Lett., 195B (2), 1987.Google Scholar
  16. [16]
    A. D. Kennedy. Hybrid Monte Carlo: In A. Billoire, R. Lacaze, A. Morel, O. Napoly, and J. Zinn-Justin, editors, Field Theory on the Lattice, pages 576–579, 1988. 1987 International Conference on Field Theory on a Lattice, Seillac, France.Google Scholar
  17. [17]
    A. D. Kennedy and Pietro Rossi: Classical Mechanics on Group Manifolds. Technical Report FSU-SCRI-89-69, UCSD 89/03, SCRI, May 1989. To appear in Nucl. Phys. B.Google Scholar
  18. [18]
    Ian T. Drummond, Simon Duane, and Ron R. Horgan: The stochastic method for numerical simulations: higher order corrections. Nucl. Phys., B220 [FS8]: 119–136, 1983.MathSciNetADSCrossRefGoogle Scholar
  19. [19]
    Simon Duane, Richard D. Kenway, Brian J. Pendleton, and Duncan Roweth: Acceleration of gauge field dynamics. Phys. Lett., B176: 143, 1986.Google Scholar
  20. [20]
    Steven Gottlieb, W. Liu, D. Toussaint, R. L. Renken, and R. L. Sugar: Hybrid-molecular-dynamics algorithms for the numerical simulation of Quantum Chromo dynamics. Phys. Rev., D35 (8): 2531–2542, April 1987.ADSGoogle Scholar
  21. [21]
    P. J. Rossky, J. D. Doll, and H. L. Friedman: Brownian dynamics as smart Monte Carlo simulation. J. Chem. Phys., 69 (10): 4628–4633, November 1978.ADSCrossRefGoogle Scholar
  22. [22]
    Steven Gottlieb, W. Liu, D. Toussaint, and R. L. Sugar: Testing an exact algorithm for simulation of fermionic QCD. Phys. Rev., D35: 261i, 1987.Google Scholar
  23. [23]
    O. Klein: Arkiv Mat Astr. Fys., 16 (5), 1922.Google Scholar
  24. [24]
    G. G. Batrouni, G. R. Katz, A. S. Kronfeld, G. P. Lepage, B. Svetitsky, and K. G. Wilson: Langevin simulations of lattice field theories. Phys. Rev., D32: 2736, 1985.ADSGoogle Scholar
  25. [25]
    A. Ukawa and M. Fukugita: Langevin simulation including dynamical quark loops. Phys. Rev. Lett., 55: 1854, 1985.ADSCrossRefGoogle Scholar
  26. [26]
    Alan Horowitz: Phys. Lett., 156B: 89, 1985.Google Scholar
  27. [27]
    Alan Horowitz: The second order Langevin equation and numerical simulations. Nucl. Phys., B280[FS18](3): 510–522, March 1987.Google Scholar
  28. [28]
    R. V. Gavai, J. Potvin, and S. Sanielevici: Numerical tests of Langevin lattice QCD. Phys. Rev., D27 (6): 1912–1921, September 1987.Google Scholar
  29. [29]
    Rajan Gupta, Gregory W. Kilcup, and Stephen R. Sharpe: Tuning the hybrid Monte Carlo algorithm. Phys. Rev., D38 (4): 1278–1287, August 1988.ADSGoogle Scholar
  30. [30]
    Akira Ukawa: Hybrid Monte Carlo algorithm as applied to dynamical Wilson fermion. In Andreas S. Kronfeld and Paul B. Mackenzie, editors, Lattice 88, pages 463–468, 1989. Proceedings of the 1988 Symposium on Lattice Field Theory, Fermilab.Google Scholar
  31. [31]
    Paul B. Mackenzie: An Improved Hybrid Monte Carlo Method. Technical Report FERMILAB-89/100-T, Fermilab, March 1989.Google Scholar
  32. [32]
    Michael Creutz: Algorithmic alternatives. In A. Billoire, R. Lacaze, A. Morel, 0. Napoly, and J. Zinn-Justin, editors, Field Theory on the Lattice, pages 547–556, 1988. 1987 International Conference on Field Theory on a Lattice, Seillac, France.Google Scholar
  33. [33]
    Michael Creutz: Global Monte Carlo algorithms for many fermion systems. Phys. Rev., D38 (4): 1228–1238, August 1988.Google Scholar
  34. [34]
    A. D. Kennedy: Higher-order Hybrid Monte Carlo. In Andreas S. Kronfeld and Paul B. Mackenzie, editors, Lattice 88, pages 457–462, 1989. Proceedings of the 1988 Symposium on Lattice Field Theory, Fermilab.Google Scholar
  35. [35]
    Massimo Campostrini and Paolo Rossi: A Comparison of Numerical Algorithms for Dynamical Fermions. Technical Report IFUP-TH 12/89, Università di Pisa, April 1989.Google Scholar
  36. [36]
    Michael Creutz and Andreas Gocksch: Higher Order Hybrid Monte Carlo Algorithms. Technical Report, Brookhaven, March 1989.Google Scholar
  37. [37]
    A. D. Kennedy and B. J. Pendleton: Some Exact Results for Hybrid Monte Carlo. Technical Report, SCRI, 1989. In preparation.Google Scholar
  38. [38]
    H. Gausterer and M. Salmhofer: Remarks on global Monte Carlo algorithms. Phys. Rev., D40 (8): 2723–2726, October 1989.ADSGoogle Scholar

Copyright information

© Springer Science+Business Media New York 1990

Authors and Affiliations

  • A. D. Kennedy
    • 1
    • 2
  1. 1.Supercomputer Computations Research Inst.Florida State UniversityTallahasseeUSA
  2. 2.Theoretical Physics DivisionFermilabBataviaUSA

Personalised recommendations