Solving Large Scale Instances of the Distribution Design Problem Using Data Mining

Abstract

In this paper we approach the solution of large instances of the distribution design problem. The traditional approaches do not consider that the instance size can significantly reduce the efficiency of the solution process. We propose a new approach that includes compression methods to transform the original instance into a new one using data mining techniques. The goal of the transformation is to condense the operation access pattern of the original instance to reduce the amount of resources needed to solve the original instance, without significantly reducing the quality of its solution. In order to validate the approach, we tested it proposing two instance compression methods on a new model of the replicated version of the distribution design problem that incorporates generalized database objects. The experimental results show that our approach permits to reduce the computational resources needed for solving large instances by at least 65%, without significantly reducing the quality of its solution. Given the encouraging results, at the moment we are working on the design and implementation of efficient instance compression methods using other data mining techniques.