Downloads & Free Reading Options - Results

Recent work concerning artificial neural networks has focused on decreasing network training times. Kernel Classifier networks, using radial basis functions (RBFs) as the kernel function, can be trained quickly with little performance degradation. Short training times are critical for systems which must adapt to changing environments. The function of Kernel Classifier networks is based on the principle that multivariate functions can be approximated via linear combinations of RBFs. RBFs can also perform probability density estimations, making classifications approximating a Baye's optimal descriminant. Methods used to set the RBF centers included matching the training data, Kohonen Training, K-Means Clustering and placement at averages of data clusters of the same class. Test results indicate the performance of these networks was equal to that of Hyperplane Classifier networks trained, via backpropagation, to optimize the Mean Square Error, Cross Entropy, and Classification Figure of Merit objective functions. However, the RBF networks trained much faster. The RBF networks also outperformed the Probability Neural Networks, (PNN) indicating the weights in the output layer offset the choice of non-optimal spreads. This ability to train quickly while obtaining high classification accuracies make RBF Kernel Classifier networks an attractive option for systems which must adapt quickly to changing environments.