Linear convergence of first order methods for non-strongly convex optimization

  • I. Necoara
  • Yu. Nesterov
  • F. Glineur
Full Length Paper Series A


The standard assumption for proving linear convergence of first order methods for smooth convex optimization is the strong convexity of the objective function, an assumption which does not hold for many practical applications. In this paper, we derive linear convergence rates of several first order methods for solving smooth non-strongly convex constrained optimization problems, i.e. involving an objective function with a Lipschitz continuous gradient that satisfies some relaxed strong convexity condition. In particular, in the case of smooth constrained convex optimization, we provide several relaxations of the strong convexity conditions and prove that they are sufficient for getting linear convergence for several first order methods such as projected gradient, fast gradient and feasible descent methods. We also provide examples of functional classes that satisfy our proposed relaxations of strong convexity conditions. Finally, we show that the proposed relaxed strong convexity conditions cover important applications ranging from solving linear systems, Linear Programming, and dual formulations of linearly constrained convex problems.

Mathematics Subject Classification

90C25 90C06 65K05 



The research leading to these results has received funding from the Executive Agency for Higher Education, Research and Innovation Funding (UEFISCDI), Romania: PN-III-P4-PCE-2016-0731, project ScaleFreeNet, No. 39/2017.


  1. 1.
    Leventhal, D., Lewis, A.S.: Randomized methods for linear constraints: convergence rates and conditioning. Math. Oper. Res. 35(3), 641–654 (2010)MathSciNetCrossRefMATHGoogle Scholar
  2. 2.
    Liu, J., Wright, S., Re, C., Bittorf, V., Sridhar, S.: An asynchronous parallel stochastic coordinate descent algorithm. J. Mach. Learn. Res. 16(1), 285–322 (2015)MathSciNetMATHGoogle Scholar
  3. 3.
    Nesterov, Y.: Introductory Lectures on Convex Optimization: A Basic Course. Kluwer, Dordrecht (2004)CrossRefMATHGoogle Scholar
  4. 4.
    Nemirovski, A., Juditsky, A., Lan, G., Shapiro, A.: Robust stochastic approximation approach to stochastic programming. SIAM J. Optim. 19(4), 1574–1609 (2009)MathSciNetCrossRefMATHGoogle Scholar
  5. 5.
    Wright, S.: Coordinate descent algorithms. Math. Program. 151(1), 3–34 (2015)MathSciNetCrossRefMATHGoogle Scholar
  6. 6.
    Burke, J.V., Deng, S.: Weak sharp minima revisited Part III: error bounds for differentiable convex inclusions. Math. Program. 116(1–2), 37–56 (2009)MathSciNetCrossRefMATHGoogle Scholar
  7. 7.
    Lewis, A.S., Pang, J.S.: Error bounds for convex inequality systems. In: Chapter In: Generalized Convexity, Generalized Monotonicity–Recent Results. Springer, Berlin (1998)Google Scholar
  8. 8.
    Yangy, T., Lin, Q.: A stochastic gradient method with linear convergence rate for a class of non-smooth non-strongly convex optimization. Tech. rep. (2015).
  9. 9.
    Luo, Z.-Q., Tseng, P.: Error bounds and convergence analysis of feasible descent methods: a general approach. Ann. Oper. Res. 46(1), 157–178 (1993)MathSciNetCrossRefMATHGoogle Scholar
  10. 10.
    Necoara, I., Clipici, D.: Parallel random coordinate descent method for composite minimization: convergence analysis and error bounds. SIAM J. Optim. 26(1), 197–226 (2016)MathSciNetCrossRefMATHGoogle Scholar
  11. 11.
    Wang, P.W., Lin, C.J.: Iteration complexity of feasible descent methods for convex optimization. J. Mach. Learn. Res. 15(4), 1523–1548 (2014)MathSciNetMATHGoogle Scholar
  12. 12.
    Zhang, H., Cheng, L.: Restricted strong convexity and its applications to convergence analysis of gradient type methods in convex optimization. Optim. Lett. 9(5), 961–979 (2015)MathSciNetCrossRefMATHGoogle Scholar
  13. 13.
    Beck, A., Shtern, S.: Linearly convergent away-step conditional gradient for non-strongly convex functions. Math. Program. 164(1–2), 1–27 (2017)MathSciNetCrossRefMATHGoogle Scholar
  14. 14.
    Drusvyatskiy, D., Lewis, A.: Error bounds, quadratic growth, and linear convergence of proximal methods. Tech. rep., (2016). (arXiv:1602.06661)
  15. 15.
    Zhou, Z., So, A.: A unified approach to error bounds for structured convex optimization problems. Math. Program. 165(2), 689–728 (2017)MathSciNetCrossRefMATHGoogle Scholar
  16. 16.
    Hoffman, A.J.: On approximate solutions of systems of linear inequalities. J. Res. Natl. Bur. Stand. 49(4), 263–265 (1952)MathSciNetCrossRefGoogle Scholar
  17. 17.
    Klatte, D., Thiere, G.: Error bounds for solutions of linear equations and inequalities. Math. Methods Oper. Res. 41(2), 191–214 (1995)MathSciNetCrossRefMATHGoogle Scholar
  18. 18.
    Nesterov, Y.: Gradient methods for minimizing composite functions. Math. Program. 140(1), 125–161 (2013)MathSciNetCrossRefMATHGoogle Scholar
  19. 19.
    O’Donoghue, B., Candes, E.: Adaptive restart for accelerated gradient schemes. Found. Comput. Math. 15(3), 715–732 (2013)MathSciNetCrossRefMATHGoogle Scholar
  20. 20.
    Bubeck, S.: Convex optimization: algorithms and complexity. Found. Trends Mach. Learn. 8(3–4), 231–357 (2015)CrossRefMATHGoogle Scholar
  21. 21.
    O’Donoghue, B., Chu, E., Parikh, N., Boyd, S.: Conic optimization via operator splitting and homogeneous self-dual embedding. J. Optim. Theory Appl. 169(3), 1042–1068 (2016)MathSciNetCrossRefMATHGoogle Scholar
  22. 22.
    Lin, H., Mairal, J., Harchaoui, Z.: A universal Catalyst for first-order optimization. In: Advances in neural information processing systems, 3384–3392 (2015)Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature and Mathematical Optimization Society 2018

Authors and Affiliations

  1. 1.Automatic Control and Systems Engineering DepartmentUniversity Politehnica BucharestBucharestRomania
  2. 2.Center for Operations Research and EconometricsUniversite catholique de LouvainLouvain-la-NeuveBelgium

Personalised recommendations