<![CDATA[Radial Basis Function Neural Network is a large group of neural network model which has distance between the input vector to the prototype vectors that are the inputs of hidden units. The benefits of Radial Basis Function Neural Network to determine the function of regulatory approaches, noisy interpolation, density estimation, optimal classification theory and potential functions. There are some benefits of Radial Basis Function Neural Network, but there is a lack of standard procedures to determine the model of Radial Basis Function Neural Networks on the optimal time series data. This study uses the data of foreign tourist visiting to Yogyakarta in 1994-2006. The data is taken from the amount of input 4 and classes of 3,4,5,6, and 7. Subsequently from them it is determined the center and variance of each class using K-Means clustering method and determined the number of basis functions on the model of Radial Basis Function Neural Network by using forward selection method. The results is that there are five types of centers based on the number of inputs and the number of classes. Based on the acquired centers, by using forward selection for the number of classes 3, 4, 5, 6, and 7 it is obtained by consecutive number of basis functions 3, 3, 4.6, and 7.]]>
Copyrights © 2011
