New Regularized Algorithms for Transductive Learning

  • Partha Pratim Talukdar
  • Koby Crammer
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5782)


We propose a new graph-based label propagation algorithm for transductive learning. Each example is associated with a vertex in an undirected graph and a weighted edge between two vertices represents similarity between the two corresponding example. We build on Adsorption, a recently proposed algorithm and analyze its properties. We then state our learning algorithm as a convex optimization problem over multi-label assignments and derive an efficient algorithm to solve this problem. We state the conditions under which our algorithm is guaranteed to converge. We provide experimental evidence on various real-world datasets demonstrating the effectiveness of our algorithm over other algorithms for such problems. We also show that our algorithm can be extended to incorporate additional prior information, and demonstrate it with classifying data where the labels are not mutually exclusive.


label propagation transductive learning graph based semi-supervised learning 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Baluja, S., Seth, R., Sivakumar, D., Jing, Y., Yagnik, J., Kumar, S., Ravichandran, D., Aly, M.: Video suggestion and discovery for youtube: taking random walks through the view graph. In: WWW (2008)Google Scholar
  2. 2.
    Belkin, M., Niyogi, P., Sindhwani, V.: Manifold regularization: A geometric framework for learning from labeled and unlabeled examples. JMLR 7, 2399–2434 (2006)MathSciNetMATHGoogle Scholar
  3. 3.
    Bengio, Y., Delalleau, O., Roux, N.: Label Propogation and Quadratic Criterion. In: Semi-Supervised Learning (2007)Google Scholar
  4. 4.
    Blitzer, J., Dredze, M., Pereira, F.: Biographies, bollywood, boom-boxes and blenders: Domain adaptation for sentiment classification. In: ACL (2007)Google Scholar
  5. 5.
    Indyk, P., Matousek, J.: Low-distortion embeddings of finite metric spaces. In: Handbook of Discrete and Computational Geometry (2004)Google Scholar
  6. 6.
    Joachims, T.: Transductive inference for text classification using support vector machines. In: ICML (1999)Google Scholar
  7. 7.
    Joachims, T.: Transductive learning via spectral graph partitioning. In: ICML (2003)Google Scholar
  8. 8.
    Katz, V.J.: The history of stokes’ theorem. Mathematics Magazine 52(3), 146–156 (1979)MathSciNetCrossRefMATHGoogle Scholar
  9. 9.
    Raghavan, V., Bollmann, P., Jung, G.: A critical investigation of recall and precision as measures of retrieval system performance. ACM TOIS 7(3), 205–229 (1989)CrossRefGoogle Scholar
  10. 10.
    Saad, Y.: Iterative Methods for Sparse Linear Systems. Society for Industrial Math. (2003)Google Scholar
  11. 11.
    Subramanya, A., Bilmes, J.: Soft-Supervised Learning for Text Classification. In: EMNLP (2008)Google Scholar
  12. 12.
    Szummer, M., Jaakkola, T.: Partially labeled classification with markov random walks. In: NIPS (2002)Google Scholar
  13. 13.
    Talukdar, P.P., Reisinger, J., Pasca, M., Ravichandran, D., Bhagat, R., Pereira, F.: Weakly supervised acquisition of labeled class instances using graph random walks. In: EMNLP (2008)Google Scholar
  14. 14.
    Wang, J., Jebara, T., Chang, S.: Graph transduction via alternating minimization. In: ICML (2008)Google Scholar
  15. 15.
    Zhu, X.: Semi-supervised learning literature survey (2005)Google Scholar
  16. 16.
    Zhu, X., Ghahramani, Z.: Learning from labeled and unlabeled data with label propagation. Technical report, CMU CALD tech report (2002)Google Scholar
  17. 17.
    Zhu, X., Ghahramani, Z., Lafferty, J.: Semi-supervised learning using gaussian fields and harmonic functions. In: ICML (2003)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Partha Pratim Talukdar
    • 1
  • Koby Crammer
    • 1
  1. 1.Computer & Information Science DepartmentUniversity of PennsylvaniaPhiladelphia

Personalised recommendations