Large-Scale Graph-Based Semi-Supervised Learning via Tree Laplacian Solver

Authors

  • Yan-Ming Zhang Institute of Automation, Chinese Academy of Sciences
  • Xu-Yao Zhang Institute of Automation, Chinese Academy of Sciences
  • Xiao-Tong Yuan Nanjing University of Information Science and Technology
  • Cheng-Lin Liu Institute of Automation, Chinese Academy of Sciences

DOI:

https://doi.org/10.1609/aaai.v30i1.10218

Keywords:

semi-supervised learning, graph-based learning methods

Abstract

Graph-based Semi-Supervised learning is one of the most popular and successful semi-supervised learning methods. Typically, it predicts the labels of unlabeled data by minimizing a quadratic objective induced by the graph, which is unfortunately a procedure of polynomial complexity in the sample size $n$. In this paper, we address this scalability issue by proposing a method that approximately solves the quadratic objective in nearly linear time. The method consists of two steps: it first approximates a graph by a minimum spanning tree, and then solves the tree-induced quadratic objective function in O(n) time which is the main contribution of this work. Extensive experiments show the significant scalability improvement over existing scalable semi-supervised learning methods.

Downloads

Published

2016-03-02

How to Cite

Zhang, Y.-M., Zhang, X.-Y., Yuan, X.-T., & Liu, C.-L. (2016). Large-Scale Graph-Based Semi-Supervised Learning via Tree Laplacian Solver. Proceedings of the AAAI Conference on Artificial Intelligence, 30(1). https://doi.org/10.1609/aaai.v30i1.10218

Issue

Section

Technical Papers: Machine Learning Methods