RESUMEN
Recursive orthogonal least squares (ROLS) is a numerically robust method for solving for the output layer weights of a radial basis function (RBF) network, and requires less computer memory than the batch alternative. In this paper, the use of ROLS is extended to selecting the centers of an RBF network. It is shown that the information available in an ROLS algorithm after network training can be used to sequentially select centers to minimize the network output error. This provides efficient methods for network reduction to achieve smaller architectures with acceptable accuracy and without retraining. Two selection methods are developed, forward and backward. The methods are illustrated in applications of RBF networks to modeling a nonlinear time series and a real multiinput-multioutput chemical process. The final network models obtained achieve acceptable accuracy with significant reductions in the number of required centers.
RESUMEN
An automatic computer imaging system for recording body surface topography has been developed on a microcomputer-based image processing system. The computer processes fringe patterns generated on the surface of the trunk and reconstructs the complete 3-dimensional form of the surface. From the topographic reconstruction, clinical parameters of scoliotic deformity such as Angle of Trunk Inclination are calculated at a number of levels from the upper thoracic to the sacral region. These multiple level measurements illustrate the change in deformity over the trunk and correspond to measurements obtained using conventional tactile devices on patients.