Linear Discriminant Analysis for Classification classifies observations by

Linear Discriminant Analysis for Classification classifies observations by

• A. Looking at how groups differ on average
• B. Drawing a line that best separates the groups
• C. Assigning each observation to the closest group
• D. Predicting a continuous value

Answer: (B)

Linear Discriminant Analysis in classification finds a linear boundary that best separates the classes. It builds a linear combination of features (a discriminant axis) and chooses a threshold so observations on one side are assigned to one class and those on the other side to the other class. This works because LDA seeks to maximize the separation between class means while keeping within-class variability small, resulting in a line (or hyperplane) that cleanly splits the groups. That’s why the statement about drawing a line that best separates the groups is the best description. It’s not about predicting a continuous value (that’s regression), and it’s not merely assigning to the closest group in the original space nor only looking at average differences—the boundary accounts for how spread each class is and where their means lie, to optimally separate them.

Linear Discriminant Analysis in classification finds a linear boundary that best separates the classes. It builds a linear combination of features (a discriminant axis) and chooses a threshold so observations on one side are assigned to one class and those on the other side to the other class. This works because LDA seeks to maximize the separation between class means while keeping within-class variability small, resulting in a line (or hyperplane) that cleanly splits the groups.

That’s why the statement about drawing a line that best separates the groups is the best description. It’s not about predicting a continuous value (that’s regression), and it’s not merely assigning to the closest group in the original space nor only looking at average differences—the boundary accounts for how spread each class is and where their means lie, to optimally separate them.