| ZHANG ZaikunRecruitment
I am looking for self-motivated students and postdocs to work with me on mathematical optimization.
Software engineers are also welcome to join me in developing optimization software.
See my announcement for more information.
Research interests
Theory, algorithms, and software of mathematical optimization, especially
derivative-free (zeroth-order) methods
randomized methods
large-scale problems
noisy problems
Selected works
Preprints
Z. Zhang, Scalable Derivative-Free Optimization Algorithms With Low-Dimensional Subspace Techniques, arXiv:2501.04536, 2025 [pdf]
Remark: This paper re-introduces a derivative-free subspace optimization framework originating from Chapter 5 of my PhD thesis supervised by Prof. Ya-xiang Yuan at the Chinese Academy of Sciences. The framework was presented at ICCOPT 2013 in Lisbon under the title “A Derivative-Free Optimization Algorithm With Low-Dimensional Subspace Techniques for Large-Scale Problems” as the second talk of Session Mon.A.15 chaired by J. Larson on July 29, 2013, which happened to be the birthday of the late Prof. M. J. D. Powell FRS. An algorithm following this framework was implemented by me in MATLAB in 2011 under the name NEWUOAs, which became the basis of the NEWUOAs code used at the Intel Corporation for chip design.
Papers
X. Chen, Y. He, and Z. Zhang, Tight Error Bounds for the Sign-Constrained Stiefel Manifold, SIAM J. Optim., 35:302–329, 2025 [pdf]
T. M. Ragonneau and Z. Zhang, PDFO: A Cross-Platform Package for Powell’s Derivative-Free Optimization Solvers, Math. Program. Comput., 16:535–559, 2024 [pdf]
S. Gratton, C. W. Royer, L. N. Vicente, and Z. Zhang, Complexity and Global Rates of Trust-Region Methods Based on Probabilistic Models, IMA J. Numer. Anal., 38:1579–1597, 2018 [pdf]
S. Gratton, C. W. Royer, L. N. Vicente, and Z. Zhang, Direct Search Based on Probabilistic Descent, SIAM J. Optim., 25:1515–1541, 2015 [pdf]
Z. Zhang, Sobolev Seminorm of Quadratic Functions With Applications to Derivative-Free Optimization, Math. Program., 146:77–96, 2014 [pdf]
Software
Z. Zhang, PRIMA: Reference Implementation for Powell’s methods with Modernization and Amelioration, a package for solving general nonlinear optimization problems without using derivatives
C. Huang, T. M. Ragonneau, and Z. Zhang, OptiProfiler, a platform for benchmarking optimization solvers
T. M. Ragonneau and Z. Zhang, COBYQA: Constrained Optimization BY Quadratic Approximation, a package for solving general nonlinear optimization problems without using derivatives
Z. Zhang, NEWUOAs, a package for solving large-scale nonlinear optimization problems without using derivatives
Remark: I implemented NEWUOAs in MATLAB when writing Chapter 5 of my PhD thesis in 2011; my implantation was ported to Modula-3 in 2016 by M. Nystroem, a Principal Engineer at the Intel Corporation, and made available in the open-source package CM3. It has been used by Intel in the design of chips, including its flagship product Atom P5900.
Contact
Address: Room 721, School of Mathematics, Sun Yat-sen University, No. 135 Xin Gang West Road, Guangzhou 510275, China
Email: zhangzaikun [&] mail.sysu.edu.cn, zaikunzhang [&] gmail.com
Remark: My former email address at The Hong Kong Polytechnic University (PolyU), zaikun.zhang@polyu.edu.hk, was deactivated by PolyU immediately after my departure from the university in January 2025. No forwarding service is provided. Hence all emails sent to that addresses are lost without known to me. Sorry if this happen to your emails.