Advertisement

Randomization in Robustness, Estimation, and Optimization

  • B. PolyakEmail author
  • P. Shcherbakov
Chapter
Part of the Systems & Control: Foundations & Applications book series (SCFA)

Abstract

This is an attempt to discuss the following question: When is a random choice better than a deterministic one? That is, if we have an original deterministic setup, is it wise to exploit randomization methods for its solution? There exist numerous situations where the positive answer is obvious; e.g., stochastic strategies in games, randomization in experiment design, randomization of inputs in identification. Another type of problems where such approach works successfully relates to treating uncertainty, see Tempo R., Calafiore G., Dabbene F., “Randomized algorithms for analysis and control of uncertain systems,” Springer, New York, 2013. We will try to focus on several research directions including optimization problems with no uncertainty and compare known deterministic methods with their stochastic counterparts such as random descent, various versions of Monte Carlo etc., for convex and global optimization. We survey some recent results in the field and ascertain that the situation can be very different.

Notes

Acknowledgements

Financial support for this work was provided by the Russian Science Foundation through project no. 16-11-10015.

References

  1. 1.
    Agmon, S.: The relaxation method for linear inequalities. Canad. J. Math. 6, 382–393 (1954)MathSciNetzbMATHCrossRefGoogle Scholar
  2. 2.
    Barmish, B.R., Lagoa, C.M.: The uniform distribution: A rigorous justification for its use in robustness analysis. Math. Control Sign. Syst. 10(3), 203–222 (1997)MathSciNetzbMATHCrossRefGoogle Scholar
  3. 3.
    Barmish, B., Polyak, B.: A new approach to open robustness problems based on probabilistic prediction formulae. In: Proc. 13th World Congress of IFAC. San Francisco, H, 1–6 (1996)Google Scholar
  4. 4.
    Barmish, B.R., Shcherbakov, P.S.: On avoiding vertexization of robustness problems: The approximate feasibility concept. In: Proc. 39th Conference on Decision and Control, Sydney, Australia (2000)Google Scholar
  5. 5.
    Barmish, B.R., Shcherbakov, P.S.: On avoiding vertexization of robustness problems: The approximate feasibility concept. IEEE Transa Autom. Control 47(5), 819–824 (2002)MathSciNetzbMATHCrossRefGoogle Scholar
  6. 6.
    Barmish, B.R., Shcherbakov, P.S., Ross, S.R., Dabbene, F.: On positivity of polynomials: The dilation integral method. IEEE Transa Autom. Control 54(5), 965–978 (2009)MathSciNetzbMATHCrossRefGoogle Scholar
  7. 7.
    Bauschke, H.H., Borwein, J.M.: On projection algorithms for solving convex feasibility problems. SIAM Review 38(3):367–426 (1996)MathSciNetzbMATHCrossRefGoogle Scholar
  8. 8.
    Ben-Tal, A., Nemirovski, A.: Robust convex optimization. Math. Oper. Res. 23(4), 769–805 (1998)MathSciNetzbMATHCrossRefGoogle Scholar
  9. 9.
    Bertsekas, D.P., Tsitsiklis, J.N.: Parallel and Distributed Computation. Prentice Hall Inc. (1989)Google Scholar
  10. 10.
    Boyd, S., El Ghaoui, L., Feron, E., Balakrishnan, V.: Linear Matrix Inequalities in Systems and Control Theory. SIAM Publ., Philadelphia (1994)zbMATHCrossRefGoogle Scholar
  11. 11.
    Calafiore, G., Campi, M.C.: Uncertain convex programs: Randomized solutions and confidence levels. Math. Prog. 201(1), 25–46 (2005)MathSciNetzbMATHCrossRefGoogle Scholar
  12. 12.
    Calafiore, G., Campi, M.: The scenario approach to robust control design. IEEE Trams. Autom. Control 45(5), 742–753 (2006)MathSciNetzbMATHCrossRefGoogle Scholar
  13. 13.
    Calafiore G., Polyak, B.: Stochastic algorithms for exact and approximate feasibility of robust LMIs. IEEE Trans. Autom. Control. 46(11), 1755–1759 (2001)MathSciNetzbMATHCrossRefGoogle Scholar
  14. 14.
    Campi, M.; Why is resorting to fate wise? A critical look at randomized algorithms in systems and control. Eur. J. Control 16(5), 419430 (2010)MathSciNetzbMATHGoogle Scholar
  15. 15.
    Censor, Y., Cegielski, A.: Projection methods: An annotated bibliography of books and reviews. Optimization: A Journal of Math. Progr. Oper. Res. 64(11), 2343–2358 (2015)Google Scholar
  16. 16.
    Censor, Y., Herman, G.T., Jiang, M.: A note on the behavior of the randomized Kaczmarz algorithm of Strohmer and Vershynin. J. Fourier Anal. Appl. 15(4), 431–436 (2009)MathSciNetzbMATHCrossRefGoogle Scholar
  17. 17.
    Censor, Y., Zenios, S.A.: Parallel Optimization: Theory, Algorithms, and Applications. New York, NY, USA: Oxford University Press; (1997)zbMATHGoogle Scholar
  18. 18.
    Dabbene, F., Scherbakov, P.S., Polyak, B.T.: A randomized cutting plane method with probabilistic geometric convergence. SIAM J. Optimiz.20(6), 3185–3207 (2010)MathSciNetzbMATHCrossRefGoogle Scholar
  19. 19.
    Dorea, C.: Expected number of steps of a random optimization method. J.Optimiz. Th. Appl. 39(2), 165–171 (1983)MathSciNetzbMATHCrossRefGoogle Scholar
  20. 20.
    Eldar, E., Needell, D.: Acceleration of randomized Kaczmarz method via the Johnson Lindenstrauss Lemma Numerical Algorithms 58(2), 163–177 (2011)MathSciNetzbMATHCrossRefGoogle Scholar
  21. 21.
    Ermoliev, Yu., Wets, R. (eds.): Numerical Techniques for Stochastic Optimization. Springer (1988)Google Scholar
  22. 22.
    Fisher, R.A.: The Design of Experiments. Oliver and Boyd, Edinburgh (1935)Google Scholar
  23. 23.
    Gilks, W.R., Richardson, S., Spiegelhalter, D.J.: Markov Chain Monte Carlo in Practice. Chapman and Hall, London (1996)zbMATHGoogle Scholar
  24. 24.
    Goldberg, D.: Genetic Algorithms in Search, Optimization and Machine Learning. Addison-Wesley, Reading, MA (1989)zbMATHGoogle Scholar
  25. 25.
    Goldensluger, A, Polyak, B.: Estimation of regression parameters with arbitrary noise. Math. Meth. Stat. 2(1), 18–29 (1993)MathSciNetGoogle Scholar
  26. 26.
    Gower, M., Richtarik, P.: Randomized iterative methods for linear systems. SIAM J. Matr. Anal. Appl. 36(4)1660–1690 (2015)MathSciNetzbMATHCrossRefGoogle Scholar
  27. 27.
    Granichin, O.: Estimating the parameters of linear regression in an arbitrary noise. Autom. Remote Control 63(1), 25–35 (2002)MathSciNetzbMATHCrossRefGoogle Scholar
  28. 28.
    Granichin, O., Polyak, B.: Randomized Algorithms for Estimation and Optimization under Almost Arbitrary Noises. Nauka, Moscow (2003) (in Russian)Google Scholar
  29. 29.
    Granichin, O., Volkovich, Z., Toledano-Kitai, D.: Randomized Algorithms in Automatic Control and Data Mining. Springer, Berlin-Heidelberg (2015)CrossRefGoogle Scholar
  30. 30.
    Gryazina, E.N., Polyak, B.: Random sampling: Billiard Walk algorithm. Eur. J. Oper. Res. 238(2), 497–504 (2014)MathSciNetzbMATHCrossRefGoogle Scholar
  31. 31.
    Gubin, L., Polyak, B., Raik, E.: The method of projections for finding the common point of convex sets. USSR Comput. Math. Math. Phys. 7(6), 1-24 (1967)CrossRefGoogle Scholar
  32. 32.
    Gupal, A.: A method for the minimization of almost-differentiable functions. Cybernetics. (1), 115–117 (1977)Google Scholar
  33. 33.
    Gürbüzbalaban, M. Ozdaglar, A., Parrilo, P.: Why random reshuffling beats stochastic gradient descent, arXiv:1510.08560v2 [math.OC], May 1, 2018.
  34. 34.
    Heunis, A.J.: Use of Monte Carlo method in an algorithm which solves a set of functional inequalities. J. Optim. Theory Appl. 45(1), 89–99 (1984)MathSciNetzbMATHCrossRefGoogle Scholar
  35. 35.
    Horst, R., Pardalos, Panos M. (eds.): Handbook of Global Optimization, vol. 1. Kluwer, Dordrecht (1995)zbMATHGoogle Scholar
  36. 36.
    Kaczmarz, S.: Angenäherte Aufslösung von Systemen linearer Gleichungen. Bull. Intern. Acad. Polon. Sci., Lett. A. 355–357 (1937). English translation: Approximate solution of systems of linear equations. Int. J. Control 57(6), 1269–1271 (1993)Google Scholar
  37. 37.
    Kroese, D.P., Taimre, T., Botev,Z.I.: Handbook of Monte Carlo Methods. John Wiley and Sons, New York (2011)zbMATHCrossRefGoogle Scholar
  38. 38.
    Kushner, H.J., Clark, D.S.: Stochastic Approximation Methods for Constrained and Unconstrained Systems. Vol. 26 of Applied Mathematical Sciences. Springer, New York (1978)zbMATHCrossRefGoogle Scholar
  39. 39.
    Lagoa, C.M., Li, X., Sznaier, M.: Probabilistically constrained linear programs and risk-adjusted controller design. SIAM J. Optimiz. 15(3), 938–951 (2005)MathSciNetzbMATHCrossRefGoogle Scholar
  40. 40.
    Lagoa, C.M., Shcherbakov, P.S., Barmish, B.R.: Probabilistic enhancement of classical robustness margins: The unirectangularity concept. Syst. Control Lett. 35(1), 31–43 (1998)MathSciNetzbMATHCrossRefGoogle Scholar
  41. 41.
    Leventhal, D., Lewis, A.S.: Randomized methods for linear constraints: Convergence rates and conditioning. Math. Oper. Res. 35(3) 641–654 (2010)MathSciNetzbMATHCrossRefGoogle Scholar
  42. 42.
    Matyas, J.: Random optimization. Autom. Remote Control 26(2), 246–253 (1965)MathSciNetzbMATHGoogle Scholar
  43. 43.
    Metropolis, N., Ulam S.: The Monte Carlo method. J. Amer. Stat. Assoc. 44(247), 335–341 (1949)zbMATHCrossRefGoogle Scholar
  44. 44.
    Motzkin, T.S., Shoenberg, I.J.: The relaxation method for linear inequalities. Canad. J. Math. 6, 393–404 (1954)MathSciNetCrossRefGoogle Scholar
  45. 45.
    Nedic, A.: Random algorithms for convex minimization problems. Math. Progr. 129(2), 225–253 (2011)MathSciNetzbMATHCrossRefGoogle Scholar
  46. 46.
    Nemirovskii, A.S, Polyak, B.T.: Necessary conditions for the stability of polynomials and their use. Autom. Remote Control 55(11), 1644–1649 (1994)zbMATHGoogle Scholar
  47. 47.
    Nemirovski, A., Yudin, D.B.: Problem Complexity and Method Efficiency in Optimization. John Wiley and Sons, New York (1983)Google Scholar
  48. 48.
    Nesterov, Yu.: Introductory Lectures on Convex Optimization: A Basic Course. Klüwer (2004)Google Scholar
  49. 49.
    Nesterov, Y.: Efficiency of coordinate descent methods on huge-scale optimization problems. SIAM J. Optimiz. 22(2), 341–362 (2012)MathSciNetzbMATHCrossRefGoogle Scholar
  50. 50.
    Nesterov, Yu., Spokoiny, V.: Random gradient-free minimization of convex functions. Foundations Comput. Math. 17(2), 527–566 (2017)MathSciNetzbMATHCrossRefGoogle Scholar
  51. 51.
    Novikova, N.M.: Stochastic quasi-gradient method for minimax seeking. USSR Comp. Math. Math. Phys. 17, 91–99 (1977)CrossRefGoogle Scholar
  52. 52.
    Pardalos, Panos M., Romeijn, H. Edwin (eds.): Handbook of Global Optimization, vol. 2. Kluwer, Dordrecht (2002)zbMATHGoogle Scholar
  53. 53.
    Petrikevich, Ya. I.: Randomized methods of stabilization of the discrete linear systems. Autom. Remote Control 69(11), 1911–1921 (2008)MathSciNetzbMATHCrossRefGoogle Scholar
  54. 54.
    Petrikevich, Ya.I., Polyak, B.T., Shcherbakov, P.S.: Fixed-order controller design for SISO systems using Monte Carlo technique. In: Proc. 9th IFAC Workshop “Adaptation and Learning in Control and Signal Processing” (ALCOSP’07) St.Petersburg, Russia (2007)CrossRefGoogle Scholar
  55. 55.
    Polyak, B.T.: Introduction to Optimization. Optimization Software, New York (1987)zbMATHGoogle Scholar
  56. 56.
    Polyak, B.: Random algorithms for solving convex inequalities. In: Butnariu, D., Censor, Y., Reich, S. (eds.) Inherently Parallel Algorithms in Feasibility and Optimization and Their Applications, pp. 409–422. Elsevier (2001)Google Scholar
  57. 57.
    Polyak, B.T., Gryazina, E.N.: Randomized methods based on new Monte Carlo schemes for control and optimization. Ann. Oper. Res. 189(1), 342–356 (2011)MathSciNetzbMATHCrossRefGoogle Scholar
  58. 58.
    Polyak, B.T., Shcherbakov, P.S.: A probabilistic approach to robust stability of time delay systems. Autom. Remote Control 57(12), 1770–1779 (1996)MathSciNetzbMATHGoogle Scholar
  59. 59.
    Polyak, B.T., Shcherbakov, P.S.: Random spherical uncertainty in estimation and robustness. IEEE Trans. Autom Control 45(11), 2145–2150 (2000)MathSciNetzbMATHCrossRefGoogle Scholar
  60. 60.
    Polyak, B., Shcherbakov, P.: Why does Monte Carlo Fail to Work Properly in High-Dimensional Optimization Problems? J. Optim. Th. Appl. 173(2), 612–627 (2017)MathSciNetzbMATHCrossRefGoogle Scholar
  61. 61.
    Polyak, B.T., Tempo, R.: Probabilistic robust design with linear quadratic regulators. Syst. Control Lett. 43(5), 343–353 (2001)MathSciNetzbMATHCrossRefGoogle Scholar
  62. 62.
    Polyak, B.T., Tsybakov, A.B.: Optimal order of accuracy for search algorithms in stochastic optimization. Problems Inform. Transmiss. 26(2), 126–133 (1990)zbMATHGoogle Scholar
  63. 63.
    Polyak, B.T., Tsybakov, A.B.: On stochastic approximation with arbitrary noise (the KW case). In: Khas’minskii, R.Z. (ed.) Topics in Nonparametric Estimation. Advances in Soviet Math. 12, 107–113 (1992)Google Scholar
  64. 64.
    Polyak, B., Yuditskij A.: Acceleration of stochastic approximation procedures by averaging. SIAM J. on Control Optimiz. 30(4), 838–855 (1992)zbMATHCrossRefGoogle Scholar
  65. 65.
    Rastrigin, L.A.: Statistical Search Method. Nauka, Moscow (1968) in Russian)Google Scholar
  66. 66.
    Richtárik, P., Tacá\(\check{\rm c}\), M.: Iteration complexity of randomized block-coordinate descent methods for minimizing a composite function. Math. Progr. 2014, 144(1–2), 1–38 (2014)MathSciNetzbMATHCrossRefGoogle Scholar
  67. 67.
    Robert, C.P., Casella, G.: Monte Carlo Statistical Methods. Springer-Verlag, New York (1999)zbMATHCrossRefGoogle Scholar
  68. 68.
    Shcherbakov, P.: Boundary oracles for control-related matrix sets. In: Proc. 19th Int. Symp. “Mathematical Theory of Networks and Systems” (MTNS-2010), Budapest, Hungary, Jul 5–9, 2010, pp. 665–670.Google Scholar
  69. 69.
    Shcherbakov, P., Dabbene, F.: On the generation of random stable polynomials. Eur. J. Control 17(2), 145–159 (2011)MathSciNetzbMATHCrossRefGoogle Scholar
  70. 70.
    Simon, D.: Evolutionary Optimization Algorithms. Wiley, New York (2013)Google Scholar
  71. 71.
    Smith, R.L.: Efficient Monte Carlo procedures for generating points uniformly distributed over bounded regions. Oper. Res. 32(6), 1296–1308 (1984)MathSciNetzbMATHCrossRefGoogle Scholar
  72. 72.
    Solis, F.J., Wets, R.J-B.: (1981). Minimization by random search techniques. Math. Oper. Res. 6(1), 19–13 (1981)MathSciNetzbMATHCrossRefGoogle Scholar
  73. 73.
    Spall, J.C.: Introduction to Stochastic Search and Optimization: Estimation, Simulation, and Control. Vol. 64 of Wiley-Interscience series in discrete mathematics and optimization. John Wiley and Sons, Hoboken, NJ (2003)Google Scholar
  74. 74.
    Stengel, R.F., Ray L.R.: Stochastic robustness of linear time-invariant control systems. IEEE Trans. Autom. Control 36(1), 82–87 (1991)MathSciNetzbMATHCrossRefGoogle Scholar
  75. 75.
    Strohmer, T., Vershynin, R.: A randomized Kaczmarz algorithm with exponential convergence. J. Fourier Anal. Appl. 15(2), 262–278 (2009)MathSciNetzbMATHCrossRefGoogle Scholar
  76. 76.
    Tempo, R., Bai, Er-Wei, Dabbene, F.: Probabilistic robustness analysis: Explicit bounds for the minimum number of samples Syst. Control Lett. 30(5), 237–242 (1997)MathSciNetzbMATHCrossRefGoogle Scholar
  77. 77.
    Tempo, R., Calafiore, G., Dabbene, F.: Randomized Algorithms for Analysis and Control of Uncertain Systems, with Applications. Springer, London (2013)zbMATHGoogle Scholar
  78. 78.
    Tremba, A., Calafiore, G., Dabbene, F., Gryazina, E., Polyak, B., Shcherbakov, P., Tempo, R.: RACT: Randomized algorithms control toolbox for MATLAB. In: Proc. 17th World Congress of IFAC, Seoul, pp. 390–395 (2008)Google Scholar
  79. 79.
    Turchin, V.F.: On the computation of multidimensional integrals by the Monte-Carlo method. Theory of Probability and its Applications, 16(4), 720–724 (1972)CrossRefGoogle Scholar
  80. 80.
    Volkov, Y.V., Zavriev, S.K.: A general stochastic outer approximation method. SIAM J. Control Optimiz. 35(4), 1387–1421 (1997)MathSciNetzbMATHCrossRefGoogle Scholar
  81. 81.
    Yakubovich, V.A.: Finite terminating algorithms for solving countable systems of inequalities and their application in problems of adaptive systems Doklady AN SSSR 189, 495–498 (1969) (in Russian)Google Scholar
  82. 82.
    Zhigljavsky, A., Z̆hilinskas, A.: Stochastic Global Optimization. Springer Science+Business Media, New York (2008)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  1. 1.Institute of Control ScienceRussian Academy of SciencesMoscowRussia

Personalised recommendations