What is the key difference between LDA and SVM when used

for classification?

a$.$ LDA assumes a linear decision boundary, while SVM can handle both linear and non$-$linear boundaries.

b$.$ Both LDA and SVM assume that the data is linearly separable.

C$.$ SVM uses class means for classification, while LDA maximizes the margin between support vectors.

d$.$ LDA finds the optimal hyperplane between classes, while SVM maximizes the margin between classes