DocumentCode :
1202074
Title :
A self-organizing map for adaptive processing of structured data
Author :
Hagenbuchner, Markus ; Sperduti, Alessandro ; Tsoi, Ah Chung
Author_Institution :
Fac. of Informatics, Univ. of Wollongong, NSW, Australia
Volume :
14
Issue :
3
fYear :
2003
fDate :
5/1/2003 12:00:00 AM
Firstpage :
491
Lastpage :
505
Abstract :
Recent developments in the area of neural networks produced models capable of dealing with structured data. Here, we propose the first fully unsupervised model, namely an extension of traditional self-organizing maps (SOMs), for the processing of labeled directed acyclic graphs (DAGs). The extension is obtained by using the unfolding procedure adopted in recurrent and recursive neural networks, with the replicated neurons in the unfolded network comprising of a full SOM. This approach enables the discovery of similarities among objects including vectors consisting of numerical data. The capabilities of the model are analyzed in detail by utilizing a relatively large data set taken from an artificial benchmark problem involving visual patterns encoded as labeled DAGs. The experimental results demonstrate clearly that the proposed model is capable of exploiting both information conveyed in the labels attached to each node of the input DAGs and information encoded in the DAG topology.
Keywords :
data structures; recurrent neural nets; self-organising feature maps; DAG; SOM; adaptive processing; artificial benchmark problem; encoded visual patterns; fully unsupervised model; labeled directed acyclic graphs; neural networks; recurrent neural networks; recursive neural networks; replicated neurons; self-organizing map; structured data; unfolded network; unfolding procedure; Artificial neural networks; Chemistry; Data mining; Data structures; Multilayer perceptrons; Neural networks; Neurons; Pattern analysis; Recurrent neural networks; Self organizing feature maps;
fLanguage :
English
Journal_Title :
Neural Networks, IEEE Transactions on
Publisher :
ieee
ISSN :
1045-9227
Type :
jour
DOI :
10.1109/TNN.2003.810735
Filename :
1199648
Link To Document :
بازگشت