A Novel Rule Ordering Approach in Classification Association Rule Mining

Abstract

Abstract.A Classification Association Rule (CAR), a common type of mined knowledge in Data Mining, describes an implicative co-occurring relationship between a set of binary-valued data-attributes (items) and a pre-defined class, expressed in the form of an "antecedent ⇒ consequent-class" rule. Classification Association Rule Mining (CARM) is a recent Classification Rule Mining (CRM) approach that builds an Association Rule Mining (ARM) based classifier using CARs. Regardless of which particular methodology is used to build it, a classifier is usually presented as an ordered CAR list, based on an applied rule ordering strategy. Five existing rule ordering mechanisms can be identified: (1) Confidence-Support-size_of_Antecedent (CSA), (2) size_of_Antecedent-ConfidenceSupport (ACS), (3) Weighted Relative Accuracy (WRA), (4) Laplace Accuracy, and (5) χ 2 Testing. In this paper, we divide the above mechanisms into two groups: (i) pure "support-confidence" framework like, and (ii) additive score assigning like. We consequently propose a hybrid rule ordering approach by combining one approach taken from (i) and another approach taken from (ii). The experimental results show that the proposed rule ordering approach performs well with respect to the accuracy of classification.

Market Basket Analysis for Mobile Showroom

“…Association rules derived depends on confidence. Frequent item set generation is done using data mining algorithms like Apriori [4], FP-Growth Algorithm [5], Eclat [6] and K-Apriori [7]. Apriori algorithm for frequent item set mining is given below.…”

Rule Selection Method in Decision Tree Models

“…그러나 그룹 별 관측치 의 개수가 균등하게 분포된 데이터일 경우 기존 분류정확도와 의 차이가 없기 때문에 이러한 문제점까지 보완하여 고안된 기 법이 가중 상대적 분류정확도(Weighted Relative Accuracy : WRACC)이며 식 (6)과 같다 (Lavrač et al, 1999). (Lavrač et al, 1999;Coenen et al, 2004;Wang et al, 2007). 식 (7)에서 는 데이터의 클래스 개 수이다.…”

Intelligent file scoring system for malware detection from the gray list

“…Hence, the highest-order rules are tested in advance and then inserted into the classifier for predicting test data objects. For rule re-ordering, there are five popular mechanisms [31]: (1) Confidence Support size of Antecedent (CSA), (2) size of Antecedent Confidence Support (ACS), (3) Weighted Relative Accuracy (WRA), (4) Laplace Accuracy, and (5) χ 2 (chi-square) measure. CSA and ACS are belong to the pure "support-confidence" framework and have been used by CBA and CMAR for rule ranking.…”

“…We use "Best First Rule" [31] method to predict the new file samples. We select the first best rule that satisfies the new file sample according to the rule list based on our hybrid CSA/χ 2 rule re-ordering method to predict whether the new case is malware or not.…”