Invited Research Talks (past five years)

1. Gradient descent with adaptive stepsize converges (nearly) linearly under fourth-order growth, NSF TRIPODS Workshop, Toyota Technological Institute at Chicago, Chicago, IL, Dec. 2024.

2. The radius of statistical efficiency, West Coast Optimization Meeting, University of British Columbia, Vancouver, BC, Canada, Sep. 2024.

3. The radius of statistical efficiency, Analysis, Geometry, and Probability Retreat, Department of Mathematics, University of Washington, Seattle, WA, Sep. 2024.

4. Optimization for large-scale learning: beyond smoothness and convexity, Mathematical Institute of Data Science Colloquium, Johns Hopkins University, Baltimore, MD, Jan. 2024.

5. Flat minima and generalization in deep learning: a case study in low rank matrix recovery, Halıcıoğlu Data Science Institute Colloquium, University of California San Diego, La Jolla, CA, Dec. 2023.

6. Unreasonably fast convergence of SGD for overparameterized problems, Optimization for Machine Learning Workshop, Conference on Neural Information Processing Systems (NeurIPS), New Orleans, LA, Dec. 2023.

7. Unreasonably fast convergence of SGD for overparameterized problems, Workshop on Modern Continuous Optimization, Massachusetts Institute of Technology, Cambridge, MA, Aug. 2023.

8. Stochastic gradient methods beyond smoothness and convexity, International Conference on Stochastic Programming, University of California Davis, Davis, CA, Jul. 2023.

9. Subgradient methods near active manifolds: saddle point avoidance, local convergence, and asymptotic normality, TILOS & OPTML Seminar, Massachusetts Institute of Technology, Cambridge, MA, Mar. 2023.

10. Flat minima generalize for low-rank matrix recovery, Information Theory and Applications Workshop, San Diego, CA, Feb. 2023.

11. Asymptotic normality in nonsmooth stochastic approximation, Workshop on Optimization and Statistical Learning, Les Houches, France, Jan. 2023.

12. Optimization algorithms beyond smoothness and convexity, Quantitative Methods Seminar, Purdue University, West Lafayette, IN, Dec. 2022.

13. Saddle points in nonsmooth optimization, Random Matrix Theory and Numerical Linear Algebra Workshop, University of Washington, Seattle, WA, Jun. 2022.

14. Flat minima generalize for low-rank matrix recovery, Deep Learning Group Seminar, University of California San Diego, La Jolla, CA, May 2022.

15. Optimization algorithms beyond smoothness and convexity, Center for Computational Mathematics Seminar, University of California San Diego, La Jolla, CA, May 2022.

16. Polyak–Łojasiewicz inequality & friends, Deep Learning Group Seminar, University of California San Diego, La Jolla, CA, May 2022.

17. Optimization algorithms beyond smoothness and convexity, Optimization and Data Science Seminar, University of California San Diego, La Jolla, CA, Apr. 2022.

18. Active strict saddles in nonsmooth optimization, Mathematics in Imaging, Data and Optimization Seminar, Rensselaer Polytechnic Institute, Troy, NY, Feb. 2022.

19. Stochastic optimization with decision-dependent distributions, Industrial & Systems Engineering Seminar, Georgia Institute of Technology, Atlanta, GA, Nov. 2021.

20. Stochastic optimization with decision-dependent distributions, Epstein Seminar, University of Southern California, Los Angeles, CA, Oct. 2021.

21. From low probability to high confidence in stochastic convex optimization, Tutte Colloquium, University of Waterloo, Waterloo, ON, Canada, Jun. 2021.

22. Stochastic optimization with decision-dependent distributions (Slides), One World Optimization Seminar (virtual), Nov. 2020.

23.  From low probability to high confidence in stochastic convex optimization (Slides), IFDS Brown-Bag  (virtual), Wisconsin-Madison, Jul. 2020.

24. Stochastic methods for nonsmooth nonconvex optimization (Slides), Electrical Engineering and Center for Statistics and Machine Learning (CSML), Princeton, NJ, Nov. 2019.

25. Robust algorithms for stochastic optimization, Machine Learning and Optimization, Microsoft Research, Redmond, WA, Aug. 2019.

26. Convergence Rates of Stochastic Algorithms in Nonsmooth Nonconvex Optimization, Mathematics of Data and Decisions (MADD), UC Davis, CA, Feb. 2019.

27. Convergence Rates of Stochastic Algorithms in Nonsmooth Nonconvex Optimization, Computational and Applied Mathematics Colloquium, U. Chicago, IL, Oct. 2018.

28. Stochastic methods for nonsmooth nonconvex optimization, Variational Analysis and Applications, Erice, Sicily, Aug. 2018.

29. Convergence rates of stochastic algorithms for nonsmooth nonconvex problems, Modern Trends in Nonconvex Optimization for Machine Learning, ICML 2018 Workshop, Stockholm, Sweden, Jul. 2018.

30. Stochastic methods for nonsmooth nonconvex optimization, Nonconvex Formulations and Algorithms in Data Sciences, U. Wisconsin Madison, WI, Aug. 2018.

31. Stochastic subgradient method converges on tame functions, ISMP 2018, Bordeaux, France, Jul. 2018.

32. Slope and geometry in variational mathematics, CNA Seminar, Carnegie Mellon, Pittsburgh, PA, Jan. 2018.

33. Algorithms for minimizing compositions of convex functions and smooth maps (Slides), CS Theory Seminar, University of Washington, Seattle, WA, Dec. 2017.

34. Structure, complexity, and conditioning in nonsmooth optimization, Mathematics colloquium, UCSD, San Diego, CA, Nov. 2017.

35. Composite nonlinear models at scale (Slides), ORIE colloquium Cornell, Ithaca, NY, Nov. 2017.

36. Efficiency of minimizing compositions of convex functions and smooth maps (Slides), Foundations of Computational Mathematics (FoCM 2017), Barcelona, Spain, July 2017.

37. Taylor-like models in nonsmooth optimization (Slides), SIAM Conf. on Optimization, Vancouver, Canada, May 2017.

38. Efficiency of inner-outer algorithms at scale (Slides), Optimization and Statistical Learning, Les Houches, France, April 2017.

39. Expanding the reach of optimal methods (conference organizer) (Slides), West Coast Optimization Meeting (WCOM 2016), University of Washington, USA, May 2016.

40. Geometry of orthogonally invariant matrix varieties (Slides), Algebra & Discrete Mathematics Seminar, UC Davis, USA, Apr. 2016.

41. Tame variational analysis, Variational analysis and applications Erice, Italy, Aug. 2015.

42. Slope and variational geometry in optimization (Slides), International Symposium on Mathematical Programming (ISMP 2015), A.W. Tucker prize session, Pittsburgh, USA, Jul. 2015.

43. The many faces of degeneracy in conic optimization (Slides), International Symposium on Mathematical Programming (ISMP 2015), Pittsburgh, USA, Jul. 2015.

44. Tame variational analysis (Slides), International conference on Variational Analysis, Optimization, and Quantitative Finance (Terry Fest), Université de Limoges, France, May. 2015.

45. Feasibility problems: from alternating projections to matrix completions (Slides), Joint meeting of the Israel Math. Union - American Math. Soc., Tel Aviv University, Israel, Jun. 2014.

46. Feasibility problems: from alternating projections to matrix completions, SIAM Conference on Optimization, San Diego, CA, May. 2014.

47. Feasibility problems: from alternating projections to matrix completions (Slides), Tutte Seminar, University of Waterloo, Waterloo, Canada, Mar. 2014.