Finding embedded network rows in linear programs I. extraction heuristics
Article Abstract:
An embedded network within a linear program is, crudely speaking, a subset of limitations that stand for conservation of flow. Three broad classes of heuristic methods for extracting large embedded networks are examined: column-scanning deletion, row-scanning addition, and row-scanning deletion A selection of implementations are presented and their performance on realistic test problems compared. Test success depends in part on several pre-processing steps that provide a scale for the constraint matrix and reserve certain columns and rows. How efficient the later network extraction is depends on the implementation, in ways which are predictable. It is more difficult to explain effectiveness, although the more expensive and sophisticated implementations seem most reliable. Much simpler implementations do sometimes find bigger networks, however.
Publication Name: Management Science
Subject: Business, general
ISSN: 0025-1909
Year: 1988
User Contributions:
Comment about this article or add new information about this topic:
Personnel tour scheduling when starting-time restrictions are present
Article Abstract:
Restricted starting-time tour-scheduling problems (RSTP) were addressed using a two-stage heuristic solution strategy. Eight heuristic procedures based on the strategy were utilized to four sets of labor requirements. Linear programming-based lower bounds to the unrestricted starting-time problem were produced and analyzed with bounds to solutions obtained using a heuristic procedure. Constraints were also utilized in the objective function associated with RSTP to limit employee ratios and the number of shift starting times. Tour schedules were determined using integer programming while the initial solution to the scheduling concern was formulated using a modified common greedy heuristic.
Publication Name: Management Science
Subject: Business, general
ISSN: 0025-1909
Year: 1998
User Contributions:
Comment about this article or add new information about this topic:
A modeling language for mathematical programming
Article Abstract:
An efficient translator is presented which utilizes as input a linear AMPL model and related data. AMPL is a new programming language developed to create computational data structures which are less prone to mistakes. The translator produces output which can be used for standard linear programming optimizers. Both the translator and AMPL admit non-complicated extensions to more basic mathematical programs that combine discrete variables or nonlinear expressions.
Publication Name: Management Science
Subject: Business, general
ISSN: 0025-1909
Year: 1990
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: Finding Certain Weakly-Efficient Vertices in Multiple Objective Linear Fractional Programming. Multiple Objective Linear Programming with Parametric Criteria Coefficients
- Abstracts: Theoretical foundation for a learning rate budget. Variance effects in cyclic production systems. Eliminating inventory in a series facility production system
- Abstracts: Optimal timing of reviews in concurrent design for manufacturability. The impact of inspection delay on process and inspection lot sizing
- Abstracts: An integer programming approach to the bandwidth packing problem. Heuristics based on spacefilling curves for combinatorial problems in Euclidean space
- Abstracts: Heuristics with constant error guarantees for the design of tree networks