To classify observations into two or more groups

To predict the group to which an observation belongs

To identify the most important predictor variables

To reduce the number of predictor variables

Linear discriminant analysis and quadratic discriminant analysis

Stepwise discriminant analysis and forward selection discriminant analysis

Canonical discriminant analysis and logistic discriminant analysis

Discriminant function analysis and cluster analysis

Linear discriminant analysis assumes that the predictor variables are normally distributed, while quadratic discriminant analysis does not.

Linear discriminant analysis is more robust to outliers than quadratic discriminant analysis.

Linear discriminant analysis is more computationally efficient than quadratic discriminant analysis.

All of the above

Select the predictor variables

Transform the data if necessary

Fit the discriminant function

Classify the observations

Evaluate the model

Marketing: To predict the likelihood that a customer will purchase a product

Finance: To predict the risk of a loan applicant defaulting on a loan

Healthcare: To predict the likelihood that a patient will develop a disease

All of the above