From association to classification: inference using weight of evidence

Association and classification are two important tasks in data analysis, machine learning, data mining and knowledge discovery. Intensive studies have been carried out in these areas recently, but how to apply discovered event associations to classification is still seldom found in current publications. We first introduce a method based on residual analysis to discover statistically significant event associations from a database. Then we propose a measure (weight of evidence) to evaluate the evidence of a significant event association in support of, or against, a certain class membership. This measure can be applied to classify an observation with respect to any attribute. With this approach, we achieve flexible prediction. Empirical results on different data sets are discussed.