A structural learning of MLP classifiers using PfSGA and its application to Korean sign language recognition

We present experimental results for a structural learning of multilayered perceptron (MLP) classifiers using PfSGA (Parameter-free Species Genetic Algorithm) and its application to the recognition of Korean sign language. The PfSGA is a combined method of the SGA (Species Genetic Algorithm) and PfSGA (Parameter-free Genetic Algorithm). The SGA is a modified GA for reducing the search space based on species concepts and PfGA is another modified GA to reduce the learning time without determining the learning parameters. Experimental results show that the proposed method could be a useful tool for choosing an appropriate architecture for high dimensions.