Customer Aggregation in Distribution Modeling,
Abstract
The issue of customer aggregation arises frequently in modeling physical distribution systems. This issue is addressed for a class of single commodity models that includes the classical transportation and capacitated facility location problems as special cases. For any proposed aggregation of customers, an a priori upper bound is given on the amount of suboptimality thereby induced in the model. This bound is of practical use because it can provide a rigorous justification of aggregations that are plausible on the basis of geographic proximity or some other natural criterion. It also suggests a novel way of using standard clustering techniques to discover customer aggregations with small associated a priori error bounds. The analytical technique used to derive these results should prove useful for obtaining similar a priori bounds for other classes of models. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1976
- Accession Number
- ADA033973
Entities
People
- Arthur M. Geoffrion
Organizations
- University of California, Los Angeles