An estimate for the loss of the classification power when a single variate is excluded from the linear discriminant function is given. A criterion is proposed for the omission of a variate without an ...

This paper presents a definitive description of neural network methodology and provides an evaluation of its advantages and disadvantages relative to statistical procedures. The development of this ...

Ordinary linear regression (OLR) assumes that response variables are continuous. Generalized Linear Models (GLMs) provide an extension to OLR since response variables can be continuous or discrete ...

Generally speaking, there are two types of outcomes (i.e. response) in statistical analysis: continuous and categorical responses. Linear Models (LM) are one of the most commonly used statistical ...