An efficient nonmonotone projected Barzilai–Borwein method for nonnegative matrix factorization with extrapolation

References

• M.V. Balashov, Maximization of a function with Lipschitz continuous gradient, J. Math. Sci. 209 (2015), pp. 12–18. doi: 10.1007/s10958-015-2482-6
   Google Scholar
• J. Barzilai and J.M. Borwein, Two-point step size gradient methods, IMA J. Numer. Anal. 8 (1988), pp. 141–148. doi: 10.1093/imanum/8.1.141
   Web of Science ®Google Scholar
• E.G. Birgin, J.M. Martínez, and M. Raydan, Nonmonotone spectral projected gradient methods on convex sets, SIAM J. Optim. 10 (2000), pp. 1196–1211. doi: 10.1137/S1052623497330963
   Web of Science ®Google Scholar
• A. Beck and M. Teboulle, A fast iterative shrinkage-thresholding algorithm for linear inverse problems, SIAM J. Imaging Sci. 2 (2009), pp. 183–202. doi: 10.1137/080716542
   Web of Science ®Google Scholar
• A. Cichocki, R. Zdunek, and S.-A. Amari, Hierarchical ALS algorithms for nonnegative matrix and 3D tensor factorization, Lect. Notes Comput. Sci. 4666 (2007), pp. 169–176. doi: 10.1007/978-3-540-74494-8_22
   Google Scholar
• Y.H. Dai, On the nonmonotone line search, J. Optim. Theory Appl. 112 (2002), pp. 315–330. doi: 10.1023/A:1013653923062
   Web of Science ®Google Scholar
• N. Gillis, The why and how of nonnegative matrix factorization, in Regularization, Optimization, Kernels, and Support Vector Machines, J.Suykens, M. Signoretto, and A. Argyriou, eds., Chapman & Hall/CRC, Machine Learning and Pattern Recognition Series, Vol. 12, 2014.
   Google Scholar
• R. Glowinski, Numerical Methods for Nonlinear Variational Problems, Springer-Verlag, New York, 1984.
   Google Scholar
• P.H. Gong and C.S. Zhang, Efficient nonnegative matrix factorization via projected Newton method, Pattern Recognit. 45 (2012), pp. 3557–3565. doi: 10.1016/j.patcog.2012.02.037
   Web of Science ®Google Scholar
• L. Grippo, F. Lampariello, and S. Lucidi, A nonmonotone linear search technique for Newton's method, SIAM J. Numer. Anal. 23 (1986), pp. 707–716. doi: 10.1137/0723046
   Web of Science ®Google Scholar
• L. Grippo and M. Sciandrone, Nonmonotone globalization techniques for the Barzilai-Borwein gradient method, Comput. Optim. Appl. 23 (2002), pp. 143–169. doi: 10.1023/A:1020587701058
   Web of Science ®Google Scholar
• N.Z. Gu and J.T. Mo, Incorporating nonmonotone strategies into the trust region method for unconstrained optimization, Comput. Math. Appl. 55 (2008), pp. 2158–2172. doi: 10.1016/j.camwa.2007.08.038
   Web of Science ®Google Scholar
• N.Y. Guan, D.C. Tao, Z.G. Luo, and B. Yuan, NeNMF: An optimal gradient method for nonnegative matrix factorization, IEEE Trans. Signal Process. 60 (2012), pp. 2882–2898. doi: 10.1109/TSP.2012.2190406
   Web of Science ®Google Scholar
• L.X. Han, M. Neumann, and U. Prasad, Alternating projected Barzilai-Borwein methods for nonnegative matrix factorization, Electron. Trans. Numer. Anal. 36 (2009), pp. 54–82.
   Web of Science ®Google Scholar
• Y.K. Huang, H.W. Liu, and S. Zhou, An efficient monotone projected Barzilai-Borwein method for nonnegative matrix factorization, Appl. Math. Lett. 45 (2015), pp. 12–17. doi: 10.1016/j.aml.2015.01.003
   Web of Science ®Google Scholar
• Y.K. Huang, H.W. Liu, and S. Zhou, Quadratic regularization projected alternating Barzilai-Borwein method for nonnegative matrix factorization, Data Mining Knowl. Discov. 29 (2015), pp. 1665–1684. doi: 10.1007/s10618-014-0390-x
   Web of Science ®Google Scholar
• D. Kim, S. Sra, I.S. Dhillon, Fast Newton-type methods for the least squares nonnegative matrix approximation problem, SIAM International Conference on Data Mining, SIAM, Philadelphia, 2007, pp. 343–354.
   Google Scholar
• D.D. Lee and H.S. Seung, Learning the parts of objects by non-negative matrix factorization, Nature401 (1999), pp. 788–791. doi: 10.1038/44565
   PubMed Web of Science ®Google Scholar
• D.D. Lee and H.S. Seung, Algorithms for non-negative matrix factorization, Adv. Neural Process. Inf. Syst. 13 (2001), pp. 556–562.
   Google Scholar
• C.J. Lin, Projected gradient methods for non-negative matrix factorization, Neural Comput. 19 (2007), pp. 2756–2779. doi: 10.1162/neco.2007.19.10.2756
   PubMed Web of Science ®Google Scholar
• Z.S. Lu and L. Xiao, On the complexity analysis of randomized block-coordinate descent methods, Math. Program. 152 (2015), pp. 615–642. doi: 10.1007/s10107-014-0800-2
   Web of Science ®Google Scholar
• Y.E. Nesterov, A method for solving the convex programming problem with convergence rate O(1/k2), Dokl. Akad. Nauk SSSR 269 (1983), pp. 543–547.
   Web of Science ®Google Scholar
• P. Paatero and U. Tapper, Positive matrix factorization: A non-negative factor model with optimal utilization of error estimates of data values, Environmetrics 5 (1994), pp. 111–126. doi: 10.1002/env.3170050203
   Web of Science ®Google Scholar
• Y.Y. Xu and W.T. Yin, A block coordinate descent method for regularized multi-convex optimization with applications to nonnegative tensor factorization and completion, SIAM J. Imaging Sci. 6 (2013), pp. 1758–1789. doi: 10.1137/120887795
   Web of Science ®Google Scholar
• Y.Y. Xu and W.T. Yin, A globally convergent algorithm for nonconvex optimization based on block coordinate update, J. Sci. Comput. 72 (2017), pp. 1–35. doi: 10.1007/s10915-017-0376-0
   Web of Science ®Google Scholar