Abstract
Learning from structured data (i.e. graphs) is a topic that has recently received the attention of the machine learning community, which proposed connectionist models such as recursive neural nets (RNN) and graph neural nets (GNN). In spite of their sound theoretical properties, RNNs and GNNs suffer some drawbacks that may limit their application. This paper outlines an alternative connectionist framework for learning discriminant functions over structured data. The approach, albeit preliminary, is simple and suitable to maximum-a-posteriori classification of broad families of graphs, and overcomes some limitations of RNNs and GNNs. The idea is to describe a graph as an algebraic relation, i.e. as a subset of the Cartesian product. The class-posterior probabilities given the relation are reduced to products of probabilistic quantities estimated using a multilayer perceptron. Experimental comparisons on tasks that were previously solved via RNNs and GNNs validate the approach.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Barabási, A.-L., Reka, A.: Emergence of scaling in random networks. Science 286, 509–512 (1999)
Bengio, Y., Simard, P., Frasconi, P.: Learning long-term dependencies with gradient descent is difficult (Special Issue on Recurrent Neural Networks, March 94). IEEE Transactions on Neural Networks 5(2), 157–166 (1994)
Bianchini, M., Maggini, M., Sarti, L.: Object recognition using multiresolution trees. In: Joint IAPR International Workshops SSPR 2006 and SPR 2006, pp. 331–339 (2006)
Bishop, C.M.: Neural Networks for Pattern Recognition. Oxford University Press, Oxford (1995)
Bollobs, B.: Random Graphs, 2nd edn. Cambridge University Press, Cambridge, UK (2001)
Caldarelli, G., Capocci, A., De Los Rios, P., Muoz, M.: Scale-free networks from varying vertex intrinsic fitness. Physical Review Letters 89(25), 258–702 (2002)
Di Massa, V., Monfardini, G., Sarti, L., Scarselli, F., Maggini, M., Gori, M.: A comparison between recursive neural networks and graph neural networks. World Congress on Computational Intelligence, pp. 778–785 (July 2006)
Duda, R.O., Hart, P.E.: Pattern Classification and Scene Analysis. Wiley, New York (1973)
Fergus, R., Perona, P., Zisserman, A.: A sparse object category model for efficient learning and exhaustive recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 380–387. IEEE Computer Society Press, Los Alamitos (2005)
Gori, M., Monfardini, G., Scarselli, F.: A new model for learning in graph domains. In: Proc. of IJCNN-05 (August 2005)
Hammer, B., Micheli, A., Sperduti, A.: Universal approximation capability of cascade correlation for structures. Neural Computation 17(5), 1109–1159 (2005)
Hammer, B., Saunders, C., Sperduti, A.: Special issue on neural networks and kernel methods for structured domains. Neural Networks 18(8), 1015–1018 (2005)
Pollack, J.: Recursive distributed representations. Artificial Intelligence 46(1–2), 77–106 (1990)
Sperduti, A., Starita, A.: Supervised neural networks for the classification of structures. IEEE Transactions on Neural Networks 8(3), 714–735 (1997)
Watts, D., Strogatz, S.: Collective dynamics of small world networks. Nature 393, 440–442 (1998)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2007 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Trentin, E., Di Iorio, E. (2007). A Simple and Effective Neural Model for the Classification of Structured Patterns. In: Apolloni, B., Howlett, R.J., Jain, L. (eds) Knowledge-Based Intelligent Information and Engineering Systems. KES 2007. Lecture Notes in Computer Science(), vol 4692. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-74819-9_2
Download citation
DOI: https://doi.org/10.1007/978-3-540-74819-9_2
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-74817-5
Online ISBN: 978-3-540-74819-9
eBook Packages: Computer ScienceComputer Science (R0)