An Improved Deterministic Rescaling for Linear Programming Algorithms

@article{Hoberg2017AnID,
  title={An Improved Deterministic Rescaling for Linear Programming Algorithms},
  author={Rebecca Hoberg and Thomas Rothvoss},
  journal={ArXiv},
  year={2017},
  volume={abs/1612.04782}
}
The perceptron algorithm for linear programming, arising from machine learning, has been around since the 1950s. While not a polynomial-time algorithm, it is useful in practice due to its simplicity and robustness. In 2004, Dunagan and Vempala showed that a randomized rescaling turns the perceptron method into a polynomial time algorithm, and later Pena and Soheili gave a deterministic rescaling. In this paper, we give a deterministic rescaling for the perceptron algorithm that improves upon… 
Geometric Rescaling Algorithms for Submodular Function Minimization
TLDR
A new class of polynomial-time algorithms for submodular function minimization (SFM) as well as a unified framework to obtain stronglyPolynomial SFM algorithms that can be applied to a wide range of combinatorial and continuous algorithms, including pseudo-polynomial ones.
Projection and Rescaling Algorithm for Finding Maximum Support Solutions to Polyhedral Conic Systems
TLDR
The algorithm is a natural but significant extension of a previous projection and rescaling algorithm that finds a solution to the full support problem: [Formula: see text] when this problem is feasible.
Rescaling Algorithms for Linear Conic Feasibility
TLDR
The degenerate case $\rho_A=0$ is addressed by extending the algorithms to find maximum support nonnegative vectors in the kernel of A and in the image of A^\top by extending them to the oracle setting.
On finding exact solutions of linear programs in the oracle model
TLDR
A general framework for extracting dual certificates based on the work of Burrell and Todd is developed and it is shown that under a natural assumption, simultaneous Diophantine approximation in these results can be avoided.
Projection and rescaling algorithm for finding most interior solutions to polyhedral conic systems
TLDR
The algorithm complements a basic procedure that involves only projections onto $L$ and $L^\perp$ with a periodic rescaling step and obtains a sharper analysis of the projection and rescaling algorithm in the latter special case.
Computational performance of a projection and rescaling algorithm
TLDR
A full description of a MATLAB implementation of a projection and rescaling algorithm for finding most interior solutions to the pair of feasibility problems and multiple sets of numerical experiments on synthetic problem instances with varied levels of conditioning provide promising evidence of the effectiveness of the algorithm.
Rescaling Algorithms for Linear Programming - Part I: Conic feasibility
TLDR
These algorithms for finding maximum support nonnegative vectors in the kernel of A and in the image of A⊤ iterate between simple first order steps and rescaling steps, improving natural geometric potentials in the domain and image spaces, respectively.

References

SHOWING 1-10 OF 40 REFERENCES
A simple polynomial-time rescaling algorithm for solving linear programs
TLDR
It is shown that a randomized version of the perceptron algorithm along with periodic rescaling runs in polynomial-time, and the resulting algorithm for linear programming has an elementary description and analysis.
A Smooth Perceptron Algorithm
TLDR
A modified version of the perceptron algorithm is proposed that retains the algorithm's original simplicity but has a substantially improved convergence rate.
A new polynomial-time algorithm for linear programming
TLDR
It is proved that given a polytopeP and a strictly interior point a εP, there is a projective transformation of the space that mapsP, a toP′, a′ having the following property: the ratio of the radius of the smallest sphere with center a′, containingP′ to theradius of the largest sphere withCenter a′ contained inP′ isO(n).
Rescaled Coordinate Descent Methods for Linear Programming
TLDR
Two simple polynomial-time algorithms to find a positive solution toAx=0 are proposed, which iterate between coordinate descent steps similar to von Neumann's algorithm, and rescaling steps, and can be extended to find the solution of maximum support for the system.
A deterministic rescaled perceptron algorithm
TLDR
A version of the perceptron algorithm that includes a periodic rescaling of the ambient space that is simpler and shorter and does not require randomization or deep separation oracles.
Approximation Algorithms
  • V. Vazirani
  • Computer Science
    Springer Berlin Heidelberg
  • 2003
Covering the basic techniques used in the latest research work, the author consolidates progress made so far, including some very recent and promising results, and conveys the beauty and excitement
Relaxation, New Combinatorial and Polynomial Algorithms for the Linear Feasibility Problem
  • U. Betke
  • Mathematics
    Discret. Comput. Geom.
  • 2004
TLDR
A new combinatorial algorithm is found that solves the homogenized linear feasibility problem of the centers of the inspheres of spherical simplices, whose facets are determined by a subset of the constraints.
A polynomial projection algorithm for linear feasibility problems
TLDR
In a polynomial number of calls to the procedure the algorithm either proves that the original system is infeasible or finds a solution in the relative interior of the feasible set.
The Design of Approximation Algorithms
TLDR
This book shows how to design approximation algorithms: efficient algorithms that find provably near-optimal solutions to discrete optimization problems.
Fast algorithms for approximate semidefinite programming using the multiplicative weights update method
TLDR
This work uses a Lagrangian-relaxation based technique to derive faster algorithms for approximately solving several families of SDP relaxations and makes improvements in approximate eigenvalue computations by using random sampling.
...
...