Exact Solutions for Recursive Principal Components Analysis of Sequences and Trees

  • Alessandro Sperduti
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4131)


We show how a family of exact solutions to the Recursive Principal Components Analysis learning problem can be computed for sequences and tree structured inputs. These solutions are derived from eigenanalysis of extended vectorial representations of the input structures and substructures. Experimental results performed on sequences and trees generated by a context-free grammar show the effectiveness of the proposed approach.


Current Input Parse Tree Input Structure Left Child Terminal Symbol 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Callan, R.E., Palmer-Brown, D. (S)RAAM: An analytical technique for fast and reliable derivation of connectionist symbol structure representations. Connection Science 9, 139–160 (1997)CrossRefGoogle Scholar
  2. 2.
    Pollack, J.B.: Recursive distributed representations. Artificial Intelligence 46, 77–105 (1990)CrossRefGoogle Scholar
  3. 3.
    Sperduti, A.: Labeling RAAM. Connection Science 6, 429–459 (1994)CrossRefGoogle Scholar
  4. 4.
    Sperduti, A., Starita, A.: Supervised neural networks for the classification of structures. IEEE Transactions on Neural Networks 8, 714–735 (1997)CrossRefGoogle Scholar
  5. 5.
    Voegtlin, T., Dominey, P.F.: Linear Recursive Distributed Representations. Neural Networks 18, 878–895 (2005)MATHCrossRefGoogle Scholar
  6. 6.
    Voegtlin, T.: Recursive Principal Components Analysis. Neural Networks 18, 1040–1050 (2005)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Alessandro Sperduti
    • 1
  1. 1.Department of Pure and Applied MathematicsUniversity of PadovaItaly

Personalised recommendations