Re: Difference between jumping emerging patterns and classification rules
Posted by:
Dang Nguyen
Date: November 28, 2017 12:57PM
Both jumping emerging patterns and classification rules are related to the supervised learning task (i.e., we consider labels in the dataset).
Assume that the dataset has only two classes 1 and -1. A jumping emerging pattern X is an itemset occurring in class 1 but not in class -1. In other words, sup(X) in class 1 is greater than 0 but sup(X) in class -1 is 0.
A classification rule is a rule where its left hand side is an itemset and its right hand side is a class, e.g., X -> 1.
Both jumping emerging patterns and classification rules demonstrate their usefulness and effectiveness in the classification task.
Re: Difference between jumping emerging patterns and classification rules
Posted by:
Dang Nguyen
Date: December 03, 2017 12:56PM
For classification rules: 1. Supervised Pattern Mining and Applications to Classification: https://link.springer.com/chapter/10.1007/978-3-319-07821-2_17
For jumping emerging patterns: 1. A survey of emerging patterns for supervised classification: https://link.springer.com/article/10.1007/s10462-012-9355-x