Classification based on the highest impact jumping emerging patterns
Tomasz Gambin, Krzysztof Walczak · 2009
In this paper, we propose a new classification algorithm based on Jumping Emerging Patterns (JEPs), that have the highest impact on classification accuracy. The core idea of our method is the application of a new ¿REAL/ALL¿ coefficient, which is used to compare the discriminating power among various groups of JEPs. The efficacy of the proposed approach was confirmed by tests performed on both synthetic and real data sets. The results show that our method may significantly improve the classification performance in comparison to other classifiers based on JEPs.