Ting Kei Pong (T. K. Pong)
 

I am currently an associate professor at Department of Applied Mathematics, the Hong Kong Polytechnic University. This is my CV.

Background

I received my Bachelor degree in 2004 from the Chinese University of Hong Kong, Department of Mathematics. I got my MPhil degree in 2006 in the same department under the supervision of Professor Kung Fu Ng. I started my PhD study in 2006 in Department of Mathematics of University of Washington, under the supervision of Professor Paul Tseng. After the disappearance of Professor Tseng, I was coadvised by Professor Maryam Fazel and Professor Rekha Thomas. I was in Simon Fraser University from July 2010 to March 2011 working with Professor Zhaosong Lu as a visiting researcher. I got my PhD degree in June 2011. From June 2011 to July 2013, I was a postdoctoral fellow at University of Waterloo, under the mentorship of Professor Stephen Vavasis and Professor Henry Wolkowicz. From July 2013 to July 2014, I was a PIMS postdoctoral fellow at University of British Columbia, working with Professor Michael Friedlander. I joined the Hong Kong Polytechnic University on Aug 1st, 2014.

Teaching

I am teaching AMA3201 and AMA615 in semester 2, 2021-22.

Current Students and Postdocs:

• Ying Lin, PhD: Chief supervisor.
• Shuqin Sun, post-doc.
• Liaoyuan Zeng, post-doc.

Former Students and Postdocs:

• Peiran Yu, PhD (2021, Chief supervisor). Current affiliation: Post-doc at University of Pittsburgh.
• Lei Yang, PhD (2017, Co-supervisor). Current affiliation: Post-doc at NUS.
• Scott B. Lindstrom, post-doc 2019-2021. Current affiliation: Post-doc at Curtin University, Australia.
• Tianxiang Liu, post-doc 2016-2018. Current affiliation: Assistant Professor at Tokyo Institute of Technology.
• Minglu Ye, post-doc 2017-2018. Current affiliation: Professor at China West Normal University.

Research Interest

Broadly speaking: Continuous optimization
Current focus:

• Convex relaxations;
• First-order methods for large-scale (convex or nonconvex) problems;

Other interests:

• Constraint qualifications for convex optimization.
• Statistical computation;
• Robust optimization.

Preprints

1. Frank-Wolfe type methods for nonconvex inequality-constrained problems (with Guoyin Li, Liaoyuan Zeng and Yongle Zhang) Submitted December 2021. Codes available at Github page hosted by Liaoyuan Zeng.
2. Tight error bounds and facial residual functions for the p-cones and beyond (with Scott B. Lindstrom and Bruno F. Lourenço) Submitted September 2021.
3. Retraction-based first-order feasible methods for difference-of-convex programs with smooth inequality and simple geometric constraints (with Guoyin Li and Yongle Zhang) Submitted June 2021. code
4. Error bounds, facial residual functions and applications to the exponential cone (with Scott B. Lindstrom and Bruno F. Lourenço) Submitted December 2020.

Publications

1. ρ-regularization subproblems: Strong duality and an eigensolver-based algorithm (with Liaoyuan Zeng) To appear in Comput. Optim. Appl. Codes available at Github page hosted by Liaoyuan Zeng.
2. Kurdyka-Lojasiewicz exponent via inf-projection (with Guoyin Li and Peiran Yu) To appear in Found. Comput. Math. DOI: https://doi.org/10.1007/s10208-021-09528-6.
3. Convergence rate analysis of a sequential convex programming method with line search for a class of constrained difference-of-convex optimization problems (with Zhaosong Lu and Peiran Yu) SIAM J. Optim. 31, 2021, pp. 2024-2054.
4. Analysis and algorithms for some compressed sensing models based on L1/L2 minimization (with Peiran Yu and Liaoyuan Zeng) SIAM J. Optim. 31, 2021, pp. 1576-1603. code
5. A strictly contractive Peaceman-Rachford splitting method for the doubly nonnegative relaxation of the minimum cut problem (with Xinxin Li, Hao Sun and Henry Wolkowicz) Comput. Optim. Appl. 78, 2021, pp. 853–891.
6. A hybrid penalty method for a class of optimization problems with multiple rank constraints (with Tianxiang Liu, Ivan Markovsky and Akiko Takeda) SIAM J. Matrix Anal. A. 41, 2020, pp. 1260-1283.
7. A difference-of-convex approach for split feasibility with applications to matrix factorizations and outlier detection (with Chen Chen, Lulin Tan and Liaoyuan Zeng) J. Global Optim. 78, 2020, pp. 107-136. See the ArXiv version for a fix of an error in Lemma 4 and Proposition 5: changed full rank to rank r in the former, and added compactness assumption to the latter. code
8. A subgradient-based approach for finding the maximum feasible subsystem with respect to a set (with Minglu Ye) SIAM J. Optim. 30, 2020, pp. 1274-1299. code
9. Inner approximating the completely positive cone via the cone of scaled diagonally dominant matrices (with Mina Saee and Joao Gouveia) J. Global Optim. 76, 2020, pp. 383-405.
10. Polar convolution (with Michael Friedlander and Ives Macedo) SIAM J. Optim. 29, 2019, pp. 1366-1391.
11. Iteratively reweighted l1 algorithms with extrapolation (with Peiran Yu) Comput. Optim. Appl. 73, 2019, pp. 353-386. code Updated on Nov 18, 2017: fixing bugs in termination criteria.
12. A refined convergence analysis of pDCAe with applications to simultaneous sparse recovery and outlier detection (with Tianxiang Liu and Akiko Takeda) Comput. Optim. Appl. 73, 2019, pp. 69-100. code
13. A nonmonotone alternating updating method for a class of matrix factorization problems (with Xiaojun Chen and Lei Yang) SIAM J. Optim. 28, 2018, pp. 3402-3430. Codes available at Lei Yang's webpage
14. A successive difference-of-convex approximation method for a class of nonconvex nonsmooth optimization problems (with Tianxiang Liu and Akiko Takeda) Math. Program. 176, 2019, pp. 339-367. DOI:10.1007/s10107-018-1327-8. code
15. A proximal difference-of-convex algorithm with extrapolation (with Xiaojun Chen and Bo Wen) Comput. Optim. Appl. 69, 2018, pp. 297-324. code.
16. Calculus of the exponent of Kurdyka-Lojasiewicz inequality and its applications to linear convergence of first-order methods (with Guoyin Li) Found. Comput. Math. 18, 2018, pp. 1199-1232. See also the ArXiv version for more details to the proof of Theorem 4.1; domains and ranges are added to the statement of Theorem 3.5 to remove ambiguity.
17. Peaceman-Rachford splitting for a class of nonconvex optimization problems (with Guoyin Li and Tianxiang Liu) Comput. Optim. Appl. 68, 2017, pp. 407-436. code
18. Two-stage stochastic variational inequalities: an ERM-solution procedure (with Xiaojun Chen and Roger Wets) Math. Program. 165, 2017, pp. 71-111.
19. Further properties of the forward-backward envelope with applications to difference-of-convex programming (with Tianxiang Liu) Comput. Optim. Appl. 67, 2017, pp. 489-520. code Codes further optimized and updated on June 26, 2016.
20. Linear convergence of proximal gradient algorithm with extrapolation for a class of nonconvex nonsmooth minimization problems (with Xiaojun Chen and Bo Wen) SIAM J. Optim. 27, 2017, pp. 124-145.
21. Alternating direction method of multipliers for a class of nonconvex and nonsmooth problems with applications to background/foreground extraction (with Xiaojun Chen and Lei Yang) SIAM J. Imaging Sci. 10, 2017, pp. 74-110.
22. Penalty methods for a class of non-Lipschitz optimization problems (with Xiaojun Chen and Zhaosong Lu) SIAM J. Optim. 26, 2016, pp. 1465-1492. code A bug in the code Lp_proj is fixed on March 6, 2017.
23. Douglas-Rachford splitting for nonconvex optimization with application to nonconvex feasibility problems (with Guoyin Li) Math. Program. 159, 2016, pp. 371-401. code
24. Eigenvalue, quadratic programming, and semidefinite programming relaxations for a cut minimization problem (with Hao Sun, Ningchuan Wang and Henry Wolkowicz) Comput. Optim. & Appl. 63, 2016, pp. 333-364. code
25. Global convergence of splitting methods for nonconvex composite optimization (with Guoyin Li) SIAM J. Optim. 25, 2015, pp. 2434-2460.
26. Gauge optimization and duality (with Michael Friedlander and Ives Macedo) SIAM J. Optim. 24, 2014 pp. 1999-2022.
27. The generalized trust region subproblem (with Henry Wolkowicz) Comput. Optim. & Appl. 58, 2014, pp. 273-322. code Note: Section 2.2.2 requires additionally b = 0. The general case is recently considered in Section 3.1 of Taati and Salahi.
28. Robust least square semidefinite programming with applications (with Guoyin Li and Alfred Ka Chun Ma) Comput. Optim. & Appl. 58, 2014, pp. 347-379. code
29. Computing optimal experimental designs via interior point method (with Zhaosong Lu) SIAM J. Matrix Anal. A. 34, 2013, pp. 1556-1580. code
30. Hankel matrix rank minimization with applications in system identification and realization (with Maryam Fazel, Defeng Sun and Paul Tseng) SIAM J. Matrix Anal. A. 34, 2013, pp. 946-977. code
31. An alternating direction method for finding Dantzig selectors (with Zhaosong Lu and Yong Zhang) Comput. Stat. Data An. 56, 2012, pp. 4037-4946. ADMDS package is available at Yong Zhang's homepage.
32. Edge-based semidefinite programming relaxation of sensor network localization with lower bound constraints Comput. Optim. & Appl. 53, 2012, pp. 23-44. code
33. Comparing SOS and SDP relaxations of sensor network localization (with Joao Gouveia) Comput. Optim. & Appl. 52, 2012, pp. 609-627. code
34. Minimizing condition number via convex programming (with Zhaosong Lu) SIAM J. Matrix Anal. A. 32, 2011, pp. 1193-1211. code
35. (Robust) Edge-based semidefinite programming relaxation of sensor network localization (with Paul Tseng) Math. Program. 130, 2011, pp. 321-358. code
36. Trace norm regularization: reformulations, algorithms, and multi-task learning (with Paul Tseng, Shuiwang Ji and Jieping Ye) SIAM J. Optim. 20, 2010, pp. 3465-3489. code
37. Constraint qualifications for convex inequality systems with applications in constrained optimization (with Chong Li and K. F. Ng) SIAM J. Optim. 19, 2008, pp. 163-187.
38. The SECQ, linear regularity, and the strong CHIP for an infinite system of closed convex sets in normed linear spaces (with Chong Li and K. F. Ng) SIAM J. Optim. 18, 2007, pp. 643-665.

Education Related Papers

1. Social resistance (with Michael Friedlander and Nathan Krislock) Comput. Sci. Eng. 18(2), 2016, pp. 98-103.

Journal Editorship

1. Associate Editor of Mathematics of Operations Research, Jan 2019 to present.

Talks

1. One World Optimization Seminar (Nov 15, 2021), Analysis and algorithms for some compressed sensing models based on the ratio of the l1 and l2 norms.
2. International Conference on Nonconvex and Distributed Optimization: Theory, Algorithm and Applications (May 29-30, 2021), Analysis and algorithms for some compressed sensing models based on the ratio of the l1 and l2 norms.
3. The 64th Annual Meeting of the Australian Mathematical Society (Dec 8-10, 2020), Convergence rate analysis of SCPls for a class of constrained difference-of-convex optimization problems.
4. The 6th International Conference on Continuous Optimization (Aug 5-8, 2019), Gauge optimization: Duality and polar envelope.
5. Tutte Seminar (June 7, 2019), Deducing Kurdyka-Lojasiewicz exponent of optimization models.
6. Mini-workshop talk at University of Tokyo (April 3, 2019), Gauge optimization: Duality and polar envelope.
7. Seminar talk at University of Tokyo (April 2, 2019), Deducing Kurdyka-Lojasiewicz exponent of optimization models.
8. The Greater Bay Area workshop on Computational Optimization 2019 (January 23 - 24, 2019), Deducing Kurdyka-Lojasiewicz exponent of optimization models.
9. EURO 2018 (July 8 - 11, 2018), Iteratively reweighted l1 algorithms with extrapolation.
10. ISMP 2018 (July 1 - 6, 2018), Iteratively reweighted l1 algorithms with extrapolation.
11. INFORMS International 2018 (Jun 17 - 20, 2018), A successive difference-of-convex approximation method for a class of nonconvex nonsmooth optimization problems.
12. SIAM Conference on Applied Linear Algebra (May 4-8, 2018), A non-monotone alternating updating method for a class of matrix factorization problems.
13. The workshop on Variational Analysis and Stochastic Optimization (Dec 11-12, 2017), Iteratively reweighted l1 algorithms with extrapolation.
14. The 5th International Conference on Continuous Optimization (Aug 6-11, 2016), Explicit estimation of KL exponent and linear convergence of 1st-order methods.
15. Trends in Optimization Seminar, UW Seattle (June 8, 2016), Explicit estimation of KL exponent and linear convergence of 1st-order methods.
16. The 5th Workshop on Optimization and Risk Management (June 2-3, 2016), Explicit estimation of KL exponent and linear convergence of 1st-order methods.
17. The 4th Workshop on Optimization and Risk Management (Dec 16-17, 2015), Splitting methods for nonconvex feasibility problems.
18. ISMP 2015 (July 12-17, 2015) Splitting methods for nonconvex feasibility problems.
19. Tutte Seminar (July 10, 2015) Splitting methods for nonconvex feasibility problems.
20. Workshop on Optimization and Data Analytics (May 13-14, 2015) Douglas-Rachford splitting for nonconvex feasibility problems.
21. The 3rd Workshop on Optimization and Risk Management (Oct 20-21, 2014), Douglas-Rachford splitting for nonconvex feasibility problems.
22. WCOM 2014 Spring (May 3, 2014), Gauge optimization and duality.
23. WCOM 2013 Autumn (Oct 5, 2013), The proximal-proximal gradient algorithm, manuscript, code
24. Optimization Days (May 6-8, 2013), Generalized trust region subproblem: analysis and algorithm.
25. UW Optimization Seminar (Nov 20, 2012), Generalized trust region subproblem.
26. ISMP 2012 (Aug 19-24, 2012), Generalized trust region subproblem: analysis and algorithm.
27. Tutte Seminar (Nov 4, 2011), Convex relaxations of sensor network localization.
28. Mid-west Optimization Meeting 2011 (Oct 14-15, 2011), Efficient solutions for large-scale trust region subproblem.
29. Thesis defense (May 10, 2011), Convex optimization in sensor network localization and multi-task learning.
30. SFU Optimization Seminar (Nov 4, 2010), SOS and SDP relaxations of sensor network localization.
31. WCOM 2010 (May 9, 2010), ESDP relaxation of sensor network localization: analysis, extensions and algorithm.
32. Talk at MIT (Nov 19, 2009), ESDP relaxation of sensor network localization: analysis, extensions and algorithm.
33. ISMP 2009 (Aug 23-28, 2009), ESDP relaxation of sensor network localization: analysis, extensions and algorithm.
34. MOPTA 2008 (Aug 18-20, 2008), ESDP relaxation of sensor network localization.
35. UW Optimization Seminar (June 3, 2008), ESDP relaxation of sensor network localization.

Links

Dr Guoyin Li at New South Wales University.
Dr Joao Gouveia at University of Coimbra.
My link on Math Genealogy project.
My link on Google Scholar.