A Hybrid Model of Fuzzy ARTMAP and the Genetic Algorithm for Data Classification

A framework for optimizing Fuzzy ARTMAP FAM neural networks using Genetic Algorithms GAs is proposed in this paper. A number of variables were identified for optimization, which include the presentation order of training data during the learning step, the feature subset selection of the training data, and the internal parameters of the FAM such as baseline vigilance and match tracking. A single configuration of all three variables were encoded as a chromosome string and evaluated by creating and training the FAM according to the variables. The fitness of the chromosome is determined by the final classification accuracy of the FAM. Evaluation on benchmark data sets are conducted with the results compared with literature. Experimental results indicate the effectiveness of the proposed framework in undertaking data classification tasks.

[1]  Yong Lu,et al.  A robust stochastic genetic algorithm (StGA) for global numerical optimization , 2004, IEEE Transactions on Evolutionary Computation.

[2]  Marek Kurzynski,et al.  A measure of competence based on random classification for dynamic ensemble selection , 2012, Inf. Fusion.

[3]  Eduardo G. Carrano,et al.  Electric distribution network multiobjective design using a problem-specific genetic algorithm , 2006, IEEE Transactions on Power Delivery.

[4]  Gail A. Carpenter,et al.  Biased ART: A neural architecture that shifts attention toward previously disregarded features following an incorrect prediction , 2010, Neural Networks.

[5]  Tung-Kuan Liu,et al.  Hybrid Taguchi-genetic algorithm for global numerical optimization , 2004, IEEE Transactions on Evolutionary Computation.

[6]  F. Pilo,et al.  A multiobjective evolutionary algorithm for the sizing and siting of distributed generation , 2005, IEEE Transactions on Power Systems.

[7]  Marzuki Khalid,et al.  Structure optimization of neural network for dynamic system modeling using multi-objective genetic algorithm , 2011, Neural Computing and Applications.

[8]  Mitsuo Gen,et al.  A Multiobjective Hybrid Genetic Algorithm for TFT-LCD Module Assembly Scheduling , 2014, IEEE Transactions on Automation Science and Engineering.

[9]  Araceli Sanchis,et al.  Genetic Approach for Optimizing Ensembles of Classifiers , 2008, FLAIRS.

[10]  Stephen Grossberg,et al.  Fuzzy ARTMAP: A neural network architecture for incremental supervised learning of analog multidimensional maps , 1992, IEEE Trans. Neural Networks.

[11]  James R. Williamson,et al.  Gaussian ARTMAP: A Neural Network for Fast Incremental Learning of Noisy Multidimensional Maps , 1996, Neural Networks.

[12]  Issam Dagher,et al.  An ordering algorithm for pattern presentation in fuzzy ARTMAP that tends to improve generalization performance , 1999, IEEE Trans. Neural Networks.

[13]  Hai Wei,et al.  A novel ensemble classifier based on multiple diverse classification methods , 2014, 2014 11th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD).

[14]  Masoud Yaghini,et al.  GOFAM: a hybrid neural network classifier combining fuzzy ARTMAP and genetic algorithm , 2011, Artificial Intelligence Review.