ABSTRACT
Prediction is performed by a quantitative target variable through the operation of the model which is typically a regular linearly regressed function. For example, based on ultrasonic scans, mechanical, chemical, and sensory attributes of beef samples can be predicted by a linear or nonlinear regression prediction model. Classification, as opposed to prediction, is performed by a categorical target variable through the model, typically, a discriminant function. For example, based on ultrasonic scans, beef samples can be classified according to USDA’s standards by a linear or nonlinear discriminant function.
