Modified ridge analyses under nonstandard conditions

A well-known procedure for the optimization of a second-degree response function over a spherical region of interest is that of ridge analysis. Khuri and Myers (1979) introduced a modification of this procedure by incorporating a certain constraint on the prediction variance. Both procedures, however, assume that the response variable has a constant variance throughout the experimental region. In the I,result article, we consider two extensions to Khuri and Myers' modified ridge analysis. The first extension relaxes the constant variance assumption. In time second extension, generalized linear models are used instead of the traditional linear models, which are commonly used in response surface methodology. The latter extension allows consideration of response variables that are not necessarily continuously distributed, including those that have discrete distributions. Two examples are presented to illustrate the implementation of the proposed extensions.