CAEP: Classification by Aggregating Emerging Patterns

Abstract

Emerging patterns (EPs) are itemsets whose supports change significantly from one dataset to another; they were recently proposed to capture multi-attribute contrasts between data classes, or trends over time. In this paper we propose a new classifier, CAEP, using the following main ideas based on EPs: (i) Each EP can sharply differentiate the class membership of a (possibly small) fraction of instances containing the EP, due to the big difference between its supports in the opposing classes; we define the differentiating power of the EP in terms of the supports and their ratio, on instances containing the EP. (ii) For each instance t, by aggregating the differentiating power of a fixed, automatically selected set of EPs, a score is obtained for each class. The scores for all classes are normalized and the largest score determines t's class. CAEP is suitable for many applications, even those with large volumes of high (e.g. 45) dimensional data; it does not depend on dimension reduction on data; and it is usually equally accurate on all classes even if their populations are unbalanced. Experiments show that CAEP has consistent good predictive accuracy, and it almost always outperforms C4.5 and CBA. By using efficient, border-based algorithms (developed elsewhere) to discover EPs, CAEP scales up on data volume and dimensionality. Observing that accuracy on the whole dataset is too coarse description of classifiers, we also used a more accurate measure, sensitivity and precision, to better characterize the performance of classifiers. CAEP is also very good under this measure.

Feature Construction Based on Closedness Properties Is Not That Simple

“…Using patterns that hold in 0/1 data as features (e.g., itemsets or association rules) is not new. Indeed, pioneering work on classification based on association rules [7] or emerging pattern discovery [8,9] have given rise to many proposals. Descriptive pattern discovery from unlabeled 0/1 data has been studied extensively during the last decade: many algorithms have been designed to compute every set pattern that satisfies a given constraint (e.g., a conjunction of constraints whose one conjunct is a minimal frequency constraint).…”

“…Several algorithms have exploited this for feature construction. Some of them select essential ones (CAEP classifier [8]) or the most expressive ones (JEPs classifier [9]). Then, an incoming example is labeled with the class c which maximizes scores based on these sets.…”

Isolated items discarding strategy for discovering high utility itemsets

“…In the future, the authors will extend the application scope of IIDS to some classification models. Classification is an important problem in data mining; several researchers have integrated classification and association rule mining [14,25]. Thus, the connection between utility mining and associative classification should be further investigated.…”

Mining discriminative items in multiple data streams

“…The applications of EPs include analyzing biological data [36,39,55] and building classifiers [21,34,35,37,38]. The central idea is to perform classification by leveraging the power of EPs present in instances to be classified.…”