Classification with Graph-Based Markov Chain

Markov chain is a popular graph-based model in data mining and machine learning areas. In this paper, we propose a novel intrinsic multi-class Markov chain classifier. It predicts the class label of arbitrary unseen data by setting the training data as the absorbing states of a Markov chain. It also incorporates kernel method for better generalization. Promising experiments on both artificial and real-world data sets demonstrate the effectiveness of our method.