Answer: 3) It maximizes the within class variance relative to the variance between classes
Step-by-step explanation:
Linear discriminant analysis (LDA), also referred to as the normal discriminant analysis (NDA), is a method that is used in machine learning, statistics, and pattern recognition, in order to get a linear combination of characteristics that can be used to separates two classes of events or more.
Of the options, the correct thing about linear discriminant analysis is that it maximizes the within class variance relative to the variance between classes.