Opportunity Cost Estimation Using Clustering and Association Rule Mining

Introduction

Data Mining is the process of extracting information from large data sets through the use of algorithms and techniques drawn from the field of statistics, machine learning and database management systems (Feelders, Daniels, & Holsheimer, 2000; Raju & Schumacker, 2016). Traditional data analysis methods often involve manual work and interpretation of data that is slow, expensive, and highly subjective. Data Mining, popularly called as knowledge discovery in large data, enables firms and organizations to make calculated decisions by assembling, accumulating, analyzing and accessing corporate data (Edvardsson, 2017). Various methods of data mining such as item-set mining, mining sequential pattern, association rule, classification, prediction, clustering, regression etc. are applied for the above areas (Han & Kamber, 2006). Further, the problem of mining association rules from transactional database was introduced by Agrawal, Imielinski, and Swami (1993). The concept aims to find frequent patterns, interesting correlations, and associations among sets of items in the transaction databases or other data repositories. Association rule are the statements that find the relationship between data in any database. Association rule has two parts “Antecedent” and “Consequent.” For example: {bread} => {eggs}. Here bread is the antecedent and egg is the consequent. Antecedent is the item that is found in the database, and consequent is the item that is found in combination with the first. Gautam & Pardasani (2010) presented an efficient version of Apriori algorithm for mining multilevel association rules in large databases for finding maximum frequent item-set at lower level of abstraction. There are two important basic measures for association rules, support and confidence. Support of an association rule is defined as the percentage of records that contain items AB to the total number of records in the database. Confidence of an association rule is defined as the percentage of the number of transactions that contain items AB to the total number of records that contain item A. The proposed algorithm can derive the multiple-level association rules under different supports in simple and effective way. To explain association rule mining, consider in a grocery store, bread is purchased with butter 30% of the time and that milk is purchased with it 40% of the time. Based on these associations, special displays of milk and bread are placed near the butter which is on sale. The management decided not to put these items on sale. These actions are aimed at increasing overall sales volume by taking advantage of frequency with which these items are purchased together. There are two association rules found in this example. The first one states that when butter is purchased, bread is purchased 30% of the time. The second one states that 40% of the time when butter is purchased so is milk. The discovered association rules can be used by management to increase the effectiveness (and reduce the cost) associated with advertising, marketing, inventory, and stock location on the floor.