A Restricted Discrete Approximation Problem in the (L sub 1) Norm.

Abstract

Discrete approximation problems in the (L sub 1) norm arise in fitting by polynomials, linear regression, and goal programming. This paper presents an algorithm to solve discrete approximations in the (L sub 1) norm when the parameters are restricted by linear constraints. The algorithm is a special purpose primal linear programming method that uses certain properties of the problem to substantially reduce storage requirements and solution times. Computational experience with a computer code version of the algorithm will be presented.

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 1975
Accession Number
ADA015608

Entities

People

  • John W. Hultz
  • Ronald D. Armstrong

Organizations

  • University of Texas at Austin

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Computer Programming
  • Computers
  • Evolutionary Algorithms
  • Goal Programming
  • Heuristic Methods
  • Linear Programming
  • Mathematics
  • Polynomials

Readers

  • Operations Research