Mining optimal item packages using mixed integer programming

Abstract

Traditional methods for discovering frequent patterns from large databases are based on attributing equal weights to all items of the database. In the real world, managerial decisions are based on economic values attached to the item sets. In this paper, we introduce the concept of the value based frequent item packages problems. Furthermore, we provide a mixed integer linear programming (MILP) model for value based optimization problem in the context of transaction data. The problem discussed in this paper is to find an optimal set of item packages (or item sets making up the whole transaction) that returns maximum profit to the organization under some limited resources. The specification of this problem opens the way for applying existing and new MILP solution techniques to deal with a number of practical decision problems. The model has been implemented and tested with real life retail data. The test results are reported in the paper.