Abstract
In this paper, a novel stochastic extra-step quasi-Newton method is developed to solve a class of nonsmooth nonconvex composite optimization problems. We assume that the gradient of the smooth part of the objective function can only be approximated by stochastic oracles. The proposed method combines general stochastic higher order steps derived from an underlying proximal type fixed-point equation with additional stochastic proximal gradient steps to guarantee convergence. Based on suitable bounds on the step sizes, we establish global convergence to stationary points in expectation and an extension of the approach using variance reduction techniques is discussed. Motivated by large-scale and big data applications, we investigate a stochastic coordinate-type quasi-Newton scheme that allows to generate cheap and tractable stochastic higher order directions. Finally, numerical results on large-scale logistic regression and deep learning problems show that our proposed algorithm compares favorably with other state-of-the-art methods.
Similar content being viewed by others
References
Agarwal, N., Bullins, B., Hazan, E.: Second-order stochastic optimization for machine learning in linear time. J. Mach. Learn. Res. 18(116), 1–40 (2017)
Akiba, T., Suzuki, S., Fukuda, K.: Extremely large minibatch SGD: training Resnet-50 on ImageNet in 15 minutes (2017). http://arxiv.org/abs/1711.04325
Allen-Zhu, Z.: Katyusha: The first direct acceleration of stochastic gradient methods. In: Proceedings of the 49th Annual ACM SIGACT Symposium on Theory of Computing, pp. 1200–1205 (2017)
Allen-Zhu, Z., Hazan, E.: Variance reduction for faster non-convex optimization. In: Proceedings of the 33rd International Conference on Machine Learning, 699–707 (2016)
Andrew, G., Gao, J.: Scalable training of \(\ell _1\)-regularized log-linear models. In: Proceedings of the 24th International Conference on Machine Learning, 33–40 (2007)
Bach, F., Jenatton, R., Mairal, J., Obozinski, G.: Optimization with sparsity-inducing penalties. Found. Trends Mach. Learn. 4(1), 1–106 (2011)
Bauschke, H.H., Combettes, P.L.: Convex analysis and monotone operator theory in Hilbert spaces. CMS books in mathematics/Ouvrages de Mathématiques de la SMC. Springer, New York (2011)
Berahas, A.S., Bollapragada, R., Nocedal, J.: An investigation of Newton-sketch and subsampled Newton methods. Optim. Methods Softw. 35(4), 661–680 (2020). https://doi.org/10.1080/10556788.2020.1725751
Berahas, A.S., Nocedal, J., Takác, M.: A multi-batch L-BFGS method for machine learning. In: Advances in Neural Information Processing Systems, pp. 1063–1071 (2016)
Bishop, C.M.: Pattern recognition and machine learning. information science and statistics. Springer, New York (2006)
Bollapragada, R., Byrd, R., Nocedal, J.: Exact and inexact subsampled Newton methods for optimization. IMA J. Numer. Anal. 39, 1–34 (2018)
Botev, A., Ritter, H., Barber, D.: Practical Gauss-Newton optimization for deep learning. In: Proceedings of the 34th International Conference on Machine Learning, 557–565 (2017)
Bottou, L., Curtis, F.E., Nocedal, J.: Optimization methods for large-scale machine learning. SIAM Rev. 60(2), 223–311 (2018)
Byrd, R.H., Chin, G.M., Neveitt, W., Nocedal, J.: On the use of stochastic Hessian information in optimization methods for machine learning. SIAM J. Optim. 21(3), 977–995 (2011)
Byrd, R.H., Hansen, S.L., Nocedal, J., Singer, Y.: A stochastic quasi-Newton method for large-scale optimization. SIAM J. Optim. 26(2), 1008–1031 (2016)
Candès, E.J., Recht, B.: Exact matrix completion via convex optimization. Found. Comput. Math. 9(6), 717–772 (2009)
Censor, Y., Gibali, A., Reich, S.: The subgradient extragradient method for solving variational inequalities in hilbert space. J. Optim. Theor. Appl. 148(2), 318–335 (2011)
Chandrasekaran, V., Sanghavi, S., Parrilo, P.A., Willsky, A.S.: Sparse and low-rank matrix decompositions. In: 27th Annual Allerton Conference on Communication, Control and Computing, 42: 1493–1498 (2009)
Chang, C.C., Lin, C.J.: LIBSVM: a library for support vector machines. ACM. Trans. Intell. Syst. Technol. 2(3), 27 (2011)
Chen, X., Qi, L.: A parameterized Newton method and a quasi-Newton method for nonsmooth equations. Comput. Optim. Appl. 3(2), 157–179 (1994)
Combettes, P.L., Pesquet, J.C.: Proximal splitting methods in signal processing. in: fixed-point algorithms for inverse problems in science and engineering. Springer, New York (2011)
Combettes, P.L., Wajs, V.R.: Signal recovery by proximal forward-backward splitting. Multiscale Model. Simul. 4(4), 1168–1200 (2005)
Conn, A.R., Gould, N.I.M., Toint, P.L.: Trust-region methods. MPS/SIAM series on optimization. SIAM. MPS, Philadelphia (2000)
Davis, D., Drusvyatskiy, D.: Stochastic subgradient method converges at the rate O\((k^{-1/4})\) on weakly convex functions (2018). http://arxiv.org/abs/1802.02988
Davis, D., Drusvyatskiy, D.: Stochastic model-based minimization of weakly convex functions. SIAM J. Optim. 29(1), 207–239 (2019). https://doi.org/10.1137/18M1178244
Defazio, A., Bach, F., Lacoste-Julien, S.: SAGA: A fast incremental gradient method with support for non-strongly convex composite objectives. In: Proceedings of the 27th International Conference on Neural Information Processing Systems, pp. 1646–1654 (2014)
Deng, L., Yu, D.: Deep learning: methods and applications. Found. Trends Signal Process. 7, 197–387 (2014)
Dong, Y.: An extension of Luque’s growth condition. Appl. Math. Lett. 22(9), 1390–1393 (2009)
Drusvyatskiy, D., Lewis, A.S.: Error bounds, quadratic growth, and linear convergence of proximal methods. Math. Oper. Res. 43(3), 919–948 (2018)
Duchi, J., Hazan, E., Singer, Y.: Adaptive subgradient methods for online learning and stochastic optimization. J. Mach. Learn. Res. 12, 2121–2159 (2011)
Durrett, R.: Probability: theory and examples, vol. 49. Cambridge University Press, Cambridge (2019)
Erdogdu, M.A., Montanari, A.: Convergence rates of sub-sampled Newton methods. In: Advances in Neural Information Processing Systems, pp. 28 (2015)
Fan, R.E., Chang, K.W., Hsieh, C.J., Wang, X.R., Lin, C.J.: LIBLINEAR: A library for large linear classification. J. Mach. Learn. Res. 9, 1871–1874 (2008)
Fang, C., Li, C.J., Lin, Z., Zhang, T.: SPIDER: Near-optimal non-convex optimization via stochastic path-integrated differential estimator. In: Proceedings of the 32nd International Conference on Neural Information Processing Systems, pp. 687–697 (2018)
Ghadimi, S., Lan, G.: Stochastic first- and zeroth-order methods for nonconvex stochastic programming. SIAM J. Optim. 23(4), 2341–2368 (2013)
Ghadimi, S., Lan, G.: Accelerated gradient methods for nonconvex nonlinear and stochastic programming. Math. Program. 156(1–2), 59–99 (2016)
Ghadimi, S., Lan, G., Zhang, H.: Mini-batch stochastic approximation methods for nonconvex stochastic composite optimization. Math. Program. 155(1–2), 267–305 (2016)
Gower, R., Goldfarb, D., Richtárik, P.: Stochastic block BFGS: Squeezing more curvature out of data. In: Proceedings of the 33rd International Conference on Machine Learning, 1869–1878 (2016)
Goyal, P., Dollár, P., Girshick, R., Noordhuis, P., Wesolowski, L., Kyrola, A., Tulloch, A., Jia, Y., He, K.: Accurate, large minibatch SGD: Training ImageNet in 1 hour (2017). http://arxiv.org/abs/1706.02677
Grosse, R., Martens, J.: A Kronecker-factored approximate Fisher matrix for convolution layers. In: Proceedings of the 33rd International Conference on Machine Learning, 573–582 (2016)
Hastie, T., Tibshirani, R., Friedman, J.: Data mining, inference, and prediction. the elements of statistical learning Springer series in statistics. Springer-Verlag, New York (2001)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778 (2016)
Hsieh, C.J., Sustik, M.A., Dhillon, I.S., Ravikumar, P.: QUIC: quadratic approximation for sparse inverse covariance estimation. J. Mach. Learn. Res. 15(1), 2911–2947 (2014)
Iusem, A.N., Jofré, A., Oliveira, R.I., Thompson, P.: Extragradient method with variance reduction for stochastic variational inequalities. SIAM J. Optim. 27(2), 686–724 (2017)
Janka, D., Kirches, C., Sager, S., Wächter, A.: An SR1/BFGS SQP algorithm for nonconvex nonlinear programs with block-diagonal Hessian matrix. Math. Program. Comput. 8(4), 435–459 (2016)
Johnson, R., Zhang, T.: Accelerating stochastic gradient descent using predictive variance reduction. Adv. in Neural Inf. Process. Syst. 26, 315–323 (2013)
Kingma, D.P., Ba, J.: Adam: A method for stochastic optimization. http://arxiv.org/abs/1412.6980 (2014)
Kohler, J.M., Lucchi, A.: Sub-sampled cubic regularization for non-convex optimization. In: Proceedings of the 34th International Conference on Machine Learning, 70:. 1895–1904 (2017)
Konečnỳ, J., Liu, J., Richtárik, P., Takáč, M.: Mini-batch semi-stochastic gradient descent in the proximal setting. IEEE J. Sel. Topics in Signal Process. 10(2), 242–255 (2016)
Korpelevich, G.: The extragradient method for finding saddle points and other problems. Matecon 12, 747–756 (1976)
LeCun, Y., Bengio, Y., Hinton, G.: Deep learning. Nature 521(7553), 436 (2015)
Lee, J.D., Sun, Y., Saunders, M.A.: Proximal Newton-type methods for minimizing composite functions. SIAM J. Optim. 24(3), 1420–1443 (2014)
Lei, L., Ju, C., Chen, J., Jordan, M.I.: Non-convex finite-sum optimization via SCSG methods. In: Proceedings of the 31st International Conference on Neural Information Processing Systems, pp. 2345–2355 (2017)
Lin, H., Mairal, J., Harchaoui, Z.: A universal catalyst for first-order optimization. In: Advances in Neural Information Processing Systems, pp. 3384–3392 (2015)
Lin, T., Ma, S., Zhang, S.: An extragradient-based alternating direction method for convex minimization. Found. Comput. Math. 17(1), 35–59 (2017)
Liu, D.C., Nocedal, J.: On the limited memory BFGS method for large scale optimization. Math. Program. 45(3), 503–528 (1989)
Liu, H., So, A.M.C., Wu, W.: Quadratic optimization with orthogonality constraint: explicit Łojasiewicz exponent and linear convergence of retraction-based line-search and stochastic variance-reduced gradient methods. Math. Program. 178, 215–262 (2018)
Liu, X., Hsieh, C.J.: Fast variance reduction method with stochastic batch size. In: Proceedings of the 35th International Conference on Machine Learning, 3185–3194 (2018)
Luo, Z.Q., Tseng, P.: Error bounds and convergence analysis of feasible descent methods: a general approach. Ann. Oper. Res. 46(1–4), 157–178 (1993)
LIBLINEAR: A library for large linear classification. http://www.csie.ntu.edu.tw/~cjlin/liblinear
Mairal, J., Bach, F., Ponce, J., Sapiro, G.: Online dictionary learning for sparse coding. In: Proceedings of the 26th International Conference on Machine Learning, 689–696 (2009)
Mannel, F., Rund, A.: A hybrid semismooth quasi-Newton method for structured nonsmooth operator equations in Banach spaces (2018). https://imsc.uni-graz.at/mannel/sqn1.pdf
Martens, J.: Deep learning via Hessian-free optimization. In: Proceedings of the 27th International Conference on Machine Learning, 27: 735–742 (2010)
Martens, J., Grosse, R.: Optimizing neural networks with Kronecker-factored approximate curvature. In: Proceedings of the 32nd International Conference on Machine Learning, pp. 2408–2417 (2015)
Mason, L., Baxter, J., Bartlett, P., Frean, M.: Boosting algorithms as gradient descent in function space. In: Proceedings of the 12th International Conference on Neural Information Processing Systems, pp. 512–518 (1999)
Milzarek, A., Xiao, X., Cen, S., Wen, Z., Ulbrich, M.: A stochastic semismooth Newton method for nonsmooth nonconvex optimization. SIAM J. Optim. 29(4), 2916–2948 (2019)
Mokhtari, A., Eisen, M., Ribeiro, A.: IQN: An incremental quasi-Newton method with local superlinear convergence rate. SIAM J. Optim. 28(2), 1670–1698 (2018)
Mokhtari, A., Ribeiro, A.: RES: regularized stochastic BFGS algorithm. IEEE Trans. Signal Process. 62(23), 6089–6104 (2014)
Mokhtari, A., Ribeiro, A.: Global convergence of online limited memory BFGS. J. Mach. Learn. Res. 16, 3151–3181 (2015)
Monteiro, R.D., Svaiter, B.F.: Complexity of variants of tseng’s modified fb splitting and korpelevich’s methods for hemivariational inequalities with applications to saddle-point and convex optimization problems. SIAM J. Optim. 21(4), 1688–1720 (2011)
Moreau, J.J.: Proximité et dualité dans un espace hilbertien. Bull. Soc. Math. FR. 93, 273–299 (1965)
Moritz, P., Nishihara, R., Jordan, M.: A linearly-convergent stochastic L-BFGS algorithm. In: Proceedings of the 19th Conference on Artificial Intelligence and Statistics, pp. 249–258 (2016)
Mutný, M.: Stochastic second-order optimization via Neumann series (2016). http://arxiv.org/abs/1612.04694
Nesterov, Y.: Gradient methods for minimizing composite functions. Math. Program. 140(1), 125–161 (2013)
Nguyen, L.M., van Dijk, M., Phan, D.T., Nguyen, P.H., Weng, T.W., Kalagnanam, J.R.: Finite-sum smooth optimization with SARAH (2019). http://arxiv.org/abs/1901.07648v2
Nguyen, L.M., Liu, J., Scheinberg, K., Takáč, M.: SARAH: A novel method for machine learning problems using stochastic recursive gradient. In: Proceedings of the 34th International Conference on Machine Learning, 2613–2621 (2017)
Nguyen, T.P., Pauwels, E., Richard, E., Suter, B.W.: Extragradient method in optimization: convergence and complexity. J. Optim. Theory Appl. 176(1), 137–162 (2018)
Nocedal, J.: Updating quasi-Newton matrices with limited storage. Math. Comp. 35(151), 773–782 (1980)
Osawa, K., Tsuji, Y., Ueno, Y., Naruse, A., Yokota, R., Matsuoka, S.: Large-scale distributed second-order optimization using Kronecker-factored approximate curvature for deep convolutional neural networks (2018). http://arxiv.org/abs/1811.12019
Pang, J.S., Qi, L.: Nonsmooth equations: motivation and algorithms. SIAM J. Optim. 3(3), 443–465 (1993)
Parikh, N., Boyd, S.: Proximal algorithms. Found. Trends Optim. 1(3), 127–239 (2014)
Paszke, A., Gross, S., Chintala, S., Chanan, G., Yang, E., DeVito, Z., Lin, Z., Desmaison, A., Antiga, L., Lerer, A.: Automatic differentiation in Pytorch. In: Proceedings of the 31th International Conference on Neural Information Processing Systems (2017)
Patrinos, P., Stella, L., Bemporad, A.: Forward-backward truncated Newton methods for convex composite optimization (2014). http://arxiv.org/abs/1402.6655
Pham, N.H., Nguyen, L.M., Phan, D.T., Tran-Dinh, Q.: ProxSARAH: an efficient algorithmic framework for stochastic composite nonconvex optimization. J. Mach. Learn. Res. 21, 1–48 (2020)
Pilanci, M., Wainwright, M.J.: Newton sketch: a near linear-time optimization algorithm with linear-quadratic convergence. SIAM J. Optim. 27(1), 205–245 (2017)
Poon, C., Liang, J., Schoenlieb, C.: Local convergence properties of SAGA/Prox-SVRG and acceleration. In: Proceedings of the 35th International Conference on Machine Learning, 80: 4124–4132 (2018)
Qi, L.: Convergence analysis of some algorithms for solving nonsmooth equations. Math. Oper. Res. 18(1), 227–244 (1993)
Qi, L.: On superlinear convergence of quasi-Newton methods for nonsmooth equations. Oper. Res. Lett. 20(5), 223–228 (1997)
Qi, L., Sun, J.: A nonsmooth version of Newton’s method. Math. Program. 58(3), 353–367 (1993)
Reddi, S.J., Hefny, A., Sra, S., Póczos, B., Smola, A.J.: Stochastic variance reduction for nonconvex optimization. In: Proceedings of the 33th International Conference on Machine Learning, 314–323 (2016)
Reddi, S.J., Sra, S., Póczos, B., Smola, A.J.: Proximal stochastic methods for nonsmooth nonconvex finite-sum optimization. In: Advances in Neural Information Processing Systems, 1145–1153 (2016)
Robbins, H., Monro, S.: A stochastic approximation method. Ann. Math. Stat. 22, 400–407 (1951)
Robbins, H., Siegmund, D.: A convergence theorem for non negative almost supermartingales and some applications. In: Optimizing Methods in Statistics, pp. 233–257. Academic Press (1971)
Rodomanov, A., Kropotov, D.: A superlinearly-convergent proximal Newton-type method for the optimization of finite sums. In: Proceeding of the 33rd International Conference on Machine Learning, 2597–2605 (2016)
Roosta-Khorasani, F., Mahoney, M.W.: Sub-sampled Newton methods. Math. Program. 76, 1–34 (2018)
Schmidhuber, J.: Deep learning in neural networks: an overview. Neural Netw. 61, 85–117 (2015)
Schmidt, M., Le Roux, N., Bach, F.: Minimizing finite sums with the stochastic average gradient. Math. Program. 162(1–2), 83–112 (2017)
Schraudolph, N.N., Yu, J., Günter, S.: A stochastic quasi-Newton method for online convex optimization. In: Proceedings of the 11th International Conference on Artificial Intelligence and Statistics, 436–443 (2007)
Shalev-Shwartz, S., Ben-David, S.: Understanding machine learning: from theory to algorithms. Cambridge University Press, Cambridge (2014)
Shalev-Shwartz, S., Tewari, A.: Stochastic methods for \(\ell _1\)-regularized loss minimization. J. Mach. Learn. Res. 12, 1865–1892 (2011)
Shalev-Shwartz, S., Zhang, T.: Stochastic dual coordinate ascent methods for regularized loss minimization. J. Mach. Learn. Res. 14, 567–599 (2013)
Shalev-Shwartz, S., Zhang, T.: Accelerated proximal stochastic dual coordinate ascent for regularized loss minimization. Math. Program. 155(1–2), 105–145 (2016)
Shi, J., Yin, W., Osher, S., Sajda, P.: A fast hybrid algorithm for large-scale \(\ell _1\)-regularized logistic regression. J. Mach. Learn. Res. 11, 713–741 (2010)
Shi, Z., Liu, R.: Large scale optimization with proximal stochastic Newton-type gradient descent In: Joint European Conference on Machine Learning and Knowledge Discovery in Databases, pp. 691–704. Springer, Cham (2015)
Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale image recognition. http://arxiv.org/abs/1409.1556 (2014)
Stella, L., Themelis, A., Patrinos, P.: Forward-backward quasi-Newton methods for nonsmooth optimization problems. Comput. Optim. Appl. 67(3), 443–487 (2017)
Sun, D., Han, J.: Newton and quasi-Newton methods for a class of nonsmooth equations and related problems. SIAM J. Optim. 7(2), 463–480 (1997)
Sutskever, I., Martens, J., Dahl, G., Hinton, G.: On the importance of initialization and momentum in deep learning. In: Proceedings of the 30th International Conference on Macine Learning, 1139–1147 (2013)
Themelis, A., Stella, L., Patrinos, P.: Forward-backward envelope for the sum of two nonconvex functions: further properties and nonmonotone linesearch algorithms. SIAM J. Optim. 28(3), 2274–2303 (2018)
Vapnik, V.: The nature of statistical learning theory. Springer Science and Business Media, New York (2013)
Wang, J., Zhang, T.: Utilizing second order information in minibatch stochastic variance reduced proximal iterations. J. Mach. Learn. Res. 20(42), 1–56 (2019)
Wang, X., Ma, C., Li, M.: A globally and superlinearly convergent quasi-Newton method for general box constrained variational inequalities without smoothing approximation. J. Global Optim. 50(4), 675–694 (2011)
Wang, X., Ma, S., Goldfarb, D., Liu, W.: Stochastic Quasi-Newton Methods for Nonconvex Stochastic Optimization. SIAM J. Optim. 27(2), 927–956 (2017)
Wang, X., Yuan, Y.X.: Stochastic proximal quasi-Newton methods for non-convex composite optimization. Optim. Methods Softw. 34, 922–948 (2019)
Wang, Z., Ji, K., Zhou, Y., Liang, Y., Tarokh, V.: Spiderboost and momentum: faster stochastic variance reduction algorithms. In: Proceedings of the 33rd International Conference on Neural Information Processing Systems (2019)
Wen, Z., Yin, W., Goldfarb, D., Zhang, Y.: A fast algorithm for sparse reconstruction based on shrinkage, subspace optimization, and continuation. SIAM J. Sci. Comput. 32(4), 1832–1857 (2010)
Xiao, L., Zhang, T.: A proximal stochastic gradient method with progressive variance reduction. SIAM J. Optim. 24(4), 2057–2075 (2014)
Xiao, X., Li, Y., Wen, Z., Zhang, L.: A regularized semi-smooth Newton method with projection steps for composite convex programs. J. Sci. Comput. 76(1), 364–389 (2018)
Xu, P., Roosta, F., Mahoney, M.W.: Newton-type methods for non-convex optimization under inexact Hessian information. Math. Program. 184, 35–70 (2019)
Xu, P., Yang, J., Roosta-Khorasani, F., Ré, C., Mahoney, M.W.: Sub-sampled Newton methods with non-uniform sampling. In: Proceedings of the 30th International Conference on Neural Information Processing Systems, pp. 3008–3016 (2016)
Xu, Y., Yin, W.: Block stochastic gradient iteration for convex and nonconvex optimization. SIAM J. Optim. 25(3), 1686–1716 (2015)
Ye, H., Luo, L., Zhang, Z.: Approximate Newton methods and their local convergence. In: Proceedings of the 34th International Conference on Machine Learning, 70: 3931–3939 (2017)
You, Y., Zhang, Z., Hsieh, C.J., Demmel, J., Keutzer, K.: ImageNet training in minutes. In: Proceedings of the 47th International Conference on Parallel Processing, 1–10 (2018)
Yuan, G.X., Ho, C.H., Lin, C.J.: An improved GLMNET for \(\ell _1\)-regularized logistic regression. J. Mach. Learn. Res. 13, 1999–2030 (2012)
Zhang, H., Reddi, S.J., Sra, S.: Riemannian SVRG: fast stochastic optimization on Riemannian manifolds. In: Advances in Neural Information Processing Systems, 4592–4600 (2016)
Zhao, R., Haskell, W.B., Tan, V.Y.: Stochastic L-BFGS: improved convergence rates and practical acceleration strategies. IEEE Trans. Signal Process 66, 1155–1169 (2017)
Zhou, D., Xu, P., Gu, Q.: Stochastic nested variance reduction for nonconvex optimization. J. Mach. Learn. Res. 21, 1–63 (2018)
Acknowledgements
The authors are grateful to the associate editor and two anonymous referees for their valuable comments and suggestions.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
M. Yang and Z. Wen are partly supported by Key-Area Research and Development Program of Guangdong Province (No. 2019B121204008), the NSFC grant 11831002 and by Beijing Academy of Artificial Intelligence.
A. Milzarek is partly supported by the Fundamental Research Fund – Shenzhen Research Institute of Big Data (SRIBD) Startup Fund JCYJ-AM20190601 and by the Shenzhen Institute of Artificial Intelligence and Robotics for Society (AIRS)
9 Appendix: proofs of auxiliary results
9 Appendix: proofs of auxiliary results
1.1 9.1 Proof of Lemma 3.2
Proof
As in [25, Lemma 3.2], applying the definition and characterization (2.3) of the proximal operator, it follows
Now, setting \(z := x + \beta d\) and \(q := \alpha d + \lambda _{+}(\nabla f(x) - \nabla f(z))\), we have \(p = \lambda _{+}[\nabla f({\bar{x}}) - \nabla f(x)] + q\) and \(\Vert q\Vert \le (\alpha + L_f \beta \lambda _{+}) \Vert d\Vert \). Furthermore, using Young’s inequality, the nonexpansiveness of the proximity operator and the Lipschitz continuity of \(\nabla f\), we obtain
where \(\mu = \alpha + L_f \beta \lambda _+\). This establishes the statement in Lemma 3.2. \(\square \)
1.2 9.2 Proof of Corollary 4.1
Proof
We need to verify that the choice of \(\lambda \) and \(\lambda _+\) satisfies the constraints derived in Theorem 4.1 and in (4.4), respectively. Notice that we can set \(\bar{\rho }= 0\) and we can work with \((\lambda _+^m)^{-1}\) instead of \((2\lambda _+^m)^{-1}\) in (4.4). Due to the definition of b and \(b_+\), we have \(\tau _m = \sqrt{2}\) for all m and thus, it follows
Furthermore, with the choice \(\lambda _+ = \gamma L^{-1}\) it holds that
and we have \({\mathcal {L}}^m_k \le \frac{1}{\lambda }\) for all k and m. Hence, the estimate (4.7) follows from (4.5).
Since each iteration in the inner loop of Algorithm 2 requires \(b+b_+\) gradient component evaluations (IFO), the total amount of evaluations in a single outer iteration is given by \(N + K(b + b_+)\) and the corresponding total number of gradient component evaluations after M outer iterations is
Moreover, since the bound for \(\mathbb {E}[\Vert F^1(\mathsf{X})\Vert ^2]\) in (4.7) is proportional to \((MK)^{-1}\), the IFO complexity of Algorithm 2 for reaching an \(\varepsilon \)-accurate stationary point with \(\mathbb {E}[\Vert F^1(\mathsf{X})\Vert ^2] \le \varepsilon \) is \({\mathcal {O}}((N/K + 2K^2)/\varepsilon )\). Minimizing this expression with respect to K yields \(K \sim N^{1/3}\) and the IFO complexity \({\mathcal {O}}(N^{2/3}/\varepsilon )\). \(\square \)
Rights and permissions
About this article
Cite this article
Yang, M., Milzarek, A., Wen, Z. et al. A stochastic extra-step quasi-Newton method for nonsmooth nonconvex optimization. Math. Program. 194, 257–303 (2022). https://doi.org/10.1007/s10107-021-01629-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10107-021-01629-y
Keywords
- Nonsmooth stochastic optimization
- Stochastic approximation
- Global convergence
- Stochastic higher order method
- Stochastic quasi-Newton scheme