A general graph-based semi-supervised learning with novel class discovery

• Published in: Neural computing & applications Vol. 19; no. 4; pp. 549 - 555
• Main Authors: Nie, Feiping, Xiang, Shiming, Liu, Yun, Zhang, Changshui
• Format: Journal Article
• Language: English
• Published: London Springer-Verlag 01.06.2010; Springer
• Subjects: Applied sciences; Artificial Intelligence; Computational Biology/Bioinformatics; Computational Science and Engineering; Computer Science; Computer science; control theory; systems; Data Mining and Knowledge Discovery; Exact sciences and technology; Image Processing and Computer Vision; Learning and adaptive systems; Mathematics; Numerical analysis; Numerical analysis in abstract spaces; Numerical analysis. Scientific computation; Numerical linear algebra; Original Article; Probability and Statistics in Computer Science; Sciences and techniques of general use; Pattern recognition; Semi-supervised learning; Normalized weights; Novel class discovery; Neural computation; Propagation; Random walk; Benchmarks; Probability distribution; Discoveries; Neural network; Interpretation; Regularization method; Algorithm performance; Supervised learning; Learning algorithm; Regularization; Latent class model
• ISSN: 0941-0643; 1433-3058
• DOI: 10.1007/s00521-009-0305-8

In this paper, we propose a general graph-based semi-supervised learning algorithm. The core idea of our algorithm is to not only achieve the goal of semi-supervised learning, but also to discover the latent novel class in the data, which may be unlabeled by the user. Based on the normalized weights evaluated on data graph, our algorithm is able to output the probabilities of data points belonging to the labeled classes or the novel class. We also give the theoretical interpretations for the algorithm from three viewpoints on graph, i.e., regularization framework, label propagation, and Markov random walks. Experiments on toy examples and several benchmark datasets illustrate the effectiveness of our algorithm.