Ting Kei Pong

Robust Least Square Semidefinite Programming with Applications to Correlation Stress Testing

  • Guoyin Li
  • Ting Kei Pong
  • K.C. Ma
    • Categories Convex and Nonsmooth Optimization, Linear, Cone and Semidefinite Programming, Robust Optimization Tags , ,

    In this paper, we consider a least square semidefinite programming problem under ellipsoidal data uncertainty. We show that the robustification of this uncertain problem can be reformulated as a semidefinite linear programming problem with an additional second-order cone constraint. We then provide an explicit quantitative sensitivity analysis on how the solution under the robustification depends … Read more

    The Generalized Trust Region Subproblem

  • Ting Kei Pong
  • Henry Wolkowicz
    • Categories Optimization Software and Modeling Systems, Unconstrained Optimization

    The \emph{interval bounded generalized trust region subproblem} (GTRS) consists in minimizing a general quadratic objective, $q_0(x) \rightarrow \min$, subject to an upper and lower bounded general quadratic constraint, $\ell \leq q_1(x) \leq u$. This means that there are no definiteness assumptions on either quadratic function. We first study characterizations of optimality for this \emph{implicitly} convex … Read more

    Hankel Matrix Rank Minimization with Applications to System Identification and Realization

  • Maryam Fazel
  • Ting Kei Pong
  • Defeng Sun
  • Paul Tseng
    • Categories Applications - Science and Engineering, Convex and Nonsmooth Optimization, Linear, Cone and Semidefinite Programming Tags , , , , ,

    We introduce a flexible optimization framework for nuclear norm minimization of matrices with linear structure, including Hankel, Toeplitz and moment structures, and catalog applications from diverse fields under this framework. We discuss various first-order methods for solving the resulting optimization problem, including alternating direction methods of multipliers, proximal point algorithms and gradient projection methods. We … Read more

    Interior Point Methods for Optimal Experimental Designs

  • Zhaosong Lu
  • Ting Kei Pong
    • Categories Constrained Nonlinear Optimization, Semi-definite Programming, Statistics Tags , , , , ,

    In this paper, we propose a primal IP method for solving the optimal experimental design problem with a large class of smooth convex optimality criteria, including A-, D- and p th mean criterion, and establish its global convergence. We also show that the Newton direction can be computed efficiently when the size of the moment … Read more

    Comparing SOS and SDP relaxations of sensor network localization

  • João Gouveia
  • Ting Kei Pong
    • Categories Semi-definite Programming Tags , , ,

    We investigate the relationships between various sum of squares (SOS) and semidefinite programming (SDP) relaxations for the sensor network localization problem. In particular, we show that Biswas and Ye’s SDP relaxation is equivalent to the degree one SOS relaxation of Kim et al. We also show that Nie’s sparse-SOS relaxation is stronger than the edge-based … Read more

    Interior Point Methods for Computing Optimal Designs

  • Zhaosong Lu
  • Ting Kei Pong
    • Categories Convex Optimization, Semi-definite Programming, Statistics Tags , , , , , ,

    In this paper we study interior point (IP) methods for solving optimal design problems. In particular, we propose a primal IP method for solving the problems with general convex optimality criteria and establish its global convergence. In addition, we reformulate the problems with A-, D- and E-criterion into linear or log-determinant semidefinite programs (SDPs) and … Read more

    Trace Norm Regularization: Reformulations, Algorithms, and Multi-task Learning

  • Shuiwang Ji
  • Ting Kei Pong
  • Paul Tseng
  • Jieping Ye
    • Categories Convex Optimization, Data-Mining, Semi-definite Programming Tags , , , , , ,

    We consider a recently proposed optimization formulation of multi-task learning based on trace norm regularized least squares. While this problem may be formulated as a semidefinite program (SDP), its size is beyond general SDP solvers. Previous solution approaches apply proximal gradient methods to solve the primal problem. We derive new primal and dual reformulations of … Read more