Aggregation of Inequalities in Integer Programming.
Abstract
Given an m x n zero-one matrix A approximately it is asked whether there is a single linear inequality ax less than or equal to b whose zero-one solutions are precisely the zero-one solutions of Ax less than or equal to e. An algorithm is developed for answering this question in O(mn sq) steps and other related problems are investigated. The results may be interpreted in terms of graph theory and threshold logic.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1975
- Accession Number
- ADA018461
Entities
People
- Peter L. Hammer
- Vaclav Chvatal
Organizations
- Stanford University