A Neuro Fuzzy Approach for Handling Structured Data

Dealing with structured data has always represented a huge problem for classical neural methods. Although many efforts have been performed, they usually pre-process data and then use classic machine learning algorithm. Another problem that machine learning algorithm have to face is the intrinsic uncertainty of data, where in such situations classic algorithm do not have the means to handle them. In this work a novel neuro-fuzzy model for structured data is presented that exploits both neural and fuzzy methods. The proposed model called Fuzzy Graph Neural Network (F-GNN) is based on GNN, a model able to handle structure data. A proof of F-GNN approximation properties is provided together with a training algorithm.

[1]  Marco Gori,et al.  A recursive neural network model for processing directed acyclic graphs with labeled edges , 2003, Proceedings of the International Joint Conference on Neural Networks, 2003..

[2]  L X Wang,et al.  Fuzzy basis functions, universal approximation, and orthogonal least-squares learning , 1992, IEEE Trans. Neural Networks.

[3]  Madan M. Gupta,et al.  Conditional Logic in Expert Systems , 1991 .

[4]  Sankar K. Pal,et al.  Multilayer perceptron, fuzzy sets, and classification , 1992, IEEE Trans. Neural Networks.

[5]  Pineda,et al.  Generalization of back-propagation to recurrent neural networks. , 1987, Physical review letters.

[6]  Jerry M. Mendel,et al.  Back-propagation fuzzy system as nonlinear dynamic system identifiers , 1992, [1992 Proceedings] IEEE International Conference on Fuzzy Systems.

[7]  Hugues Bersini,et al.  Recurrent fuzzy systems , 1994, Proceedings of 1994 IEEE 3rd International Fuzzy Systems Conference.

[8]  Ah Chung Tsoi,et al.  The Graph Neural Network Model , 2009, IEEE Transactions on Neural Networks.

[9]  Ronald J. Williams,et al.  A Learning Algorithm for Continually Running Fully Recurrent Neural Networks , 1989, Neural Computation.

[10]  P J Webros.  BACKPROPAGATION THROUGH TIME: WHAT IT DOES AND HOW TO DO IT , 1990 .

[11]  R. Challoo,et al.  A fuzzy neural hybrid system , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[12]  L. Zadeh.  Fuzzy sets as a basis for a theory of possibility , 1999 .

[13]  Chin-Teng Lin,et al.  Real-time supervised structure/parameter learning for fuzzy neural network , 1992, [1992 Proceedings] IEEE International Conference on Fuzzy Systems.

[14]  Ah Chung Tsoi,et al.  Computational Capabilities of Graph Neural Networks , 2009, IEEE Transactions on Neural Networks.

[15]  Steven J. Nowlan,et al.  Gain Variation in Recurrent Error Propagation Networks , 1988, Complex Syst..

[16]  Franco Scarselli,et al.  Recursive neural networks for processing graphs with labelled edges: theory and applications , 2005, Neural Networks.

[17]  C. S. George Lee,et al.  Neural fuzzy systems: a neuro-fuzzy synergism to intelligent systems , 1996 .

[18]  Alfredo Petrosino,et al.  Encoding nondeterministic fuzzy tree automata into recursive neural networks , 2004, IEEE Transactions on Neural Networks.