ABSTRACT
This case study has two goals. One is to show the application of the radial basis function (RBF) neural network in aiding in all aspects of design and manufacturing of advanced ceramics, where it is desirable to find which of the many processing variables contribute most to the desired properties of the material. The second goal of the chapter is to compare the RBF network results with those obtained by using fuzzy sets on the same data collected at the NASA Lewis Research Center. To set the RBF hidden layer centers and to train the output layer weights the nodes at data points and the gradient descent methods were used, respectively. The RBF network predicted strength with an average error of less than 12% and density with an average error of less than 2%, and demonstrated a potential for accelerating the development and processing of emerging ceramic materials.
